Electroanalysis for Quantitative Assessment ofBacterial Activity

著者 石木 健吾内容記述 学位記番号：論工第1578号, 指導教員：井上　博史URL http://doi.org/10.24729/00016954

Electroanalysis for Quantitative Assessment

of Bacterial Activity

（細菌活性の定量的評価のための電気化学分析）

Kengo Ishiki

石 木 健 吾

January 2020

Doctoral Thesis at Osaka Prefecture University

i

CONTENTS

CHAPTER I. Introduction 1

1.1. Bacteria 1

1.2. Color-based analysis of metabolic activity 2

1.2.1. Colorimetric assay 2

1.2.2. Fluorometric assay 3

1.2.3. Luminometric assay 4

1.3. Electrochemical techniques in bioanalysis 5

1.3.1. Voltammetric technique 5

1.3.2. Electrochemical microsystem techniques 7

1.4. Outline of the thesis 7

REFERENCES 9

CHAPTER II. A Microbial Platform Based on Conducting Polymers for Evaluating Metabolic Activity 13

2.1. Introduction 13

2.2. Experimental 14

2.2.1. Chemicals and materials 14

2.2.2. Bacterial culturing 14

2.2.3. PEDOT film fabrication 15

2.2.4. PPy film fabrication 15

2.2.5. Spectroscopy 15

2.2.6. Microscopic observation 16

2.2.7. Electrochemical measurements 16

2.3. Results and Discussion 17

2.4. Conclusion 27

REFERENCES 28

ii

CHAPTER III. Electrochemical Detection of Viable Bacterial Cells Using a Tetrazolium 31

3.1. Introduction 31

3.2. Experimental 33

3.2.1. Microbe culture and chemicals 33

3.2.2. Procedures and apparatus 34

3.3. Results and Discussion 35

3.4. Conclusion 46

REFERENCES 47

CHAPTER IV. Precious Metal-ion Reduction by Shewanella oneidensis MR-1 51

4.1. Introduction 51

4.2. Experimental 52

4.2.1. Bacterial culture and purification 52

4.2.2. Metal-ion reduction 52

4.2.3. Apparatus 52

4.2.4. Dark-field observation and measurement of light-scattering spectra 53

4.3. Results and Discussion 53

4.4. Conclusion 59

REFERENCES 60

iii

CHAPTER V. Kinetics of Intracellular Electron Generation in Shewanella oneidensis MR-1 63

5.1. Introduction 63

5.2. Experimental 65

5.2.1. Bacterial cultivation 65

5.2.2. Potentiometry in bacterial suspensions 65

5.3. Results and Discussion 66

5.4. Conclusion 77

REFERENCES 78

CHAPTER V. Summary 81

ACKOWLEDGEMENTS 84

LIST OF PUBLICATIONS 85

1

Chapter I

Introduction

1.1. Bacteria

Bacteria have existed from early in the history of life on Earth, and the number of

cells is estimated to approximately 5.0 × 1030 cells.1 Despite simplicity of its structure,

bacteria are widely found everywhere, even to the bottom of the deepest oceans2,

fumarole of volcano3, and beneath Antarctic ice sheet4. To survive such harsh

environments, bacteria expresses various functions to control the flux of electrons, ions,

and molecules in intracellular/extracellular environment. It is well-known that bacterial

activity such as cycling of nutrient and decomposing the detritus greatly contribute to

conserve ecological system. However, some bacteria have pathogenicity that cause

diseases damage to host cell, and pose serious threats such as food poisoning and infection

disease to human health. To detect such bacteria, culture methods have been widely used

as one of the most reliable techniques.5-7 These techniques allow the detection of a single

bacterium. However, the main disadvantages are multistep assay that consists of pre-

enrichment, selective enrichment, isolation, and identification steps. Completion of all

these phases requires at least 16 h but can take as long as 48 h. To circumvent this problem,

fast, simple, and sensitive techniques of bacterial detection based on metabolism,

antibody, and metal nanoparticles have been developed.8-10

On the other hand, most of bacteria lives in harmony with the human beings, and are

indispensable to our health life. Soil bacteria decompose organic matter and to provide

vital nutrients such as nitrogen and phosphorus compounds for plants. Bacterial

fermentation is utilized to produce many foods and beverages (yogurt, natto, sake, etc.).

The gut bacteria play important role in digestive health and immune systems.11,12

Furthermore, metal-reducing bacteria is expected to be applied for environmental and

energy-creation biotechnologies, including for the sewage purification, collection of

precious metal ions, and biocatalyst in microbial fuel cells.13-15 In order to utilize these

beneficial bacteria efficiently, it is essential to not only deepen our understanding of

2

bacterial functions but to evaluate them quantitatively.

1.2. Color-based analysis of metabolic activity

The conventional method for assessing metabolic activity is colony-counting that

make it possible to evaluate only viable microbial cell.16-18 However, this technique is

costly, time-consuming and laborious techniques and therefore, not suitable for a real-

time evaluation of the microbial activity. Various biological assays have been developed

to quantify metabolic activity including proliferation, respiration, viability, enzyme

activity and electron transfer. Most technique is based on colorimetric, fluorometric, or

luminometric property of redox dye.

1.2.1. Colorimetric assay

Colorimetric assay has been widely used for assessing cell activity, because color

changes can be easily recognized and applied for high-throughput screening. Tetrazolium

salts is one of the most useful tools for quantitative evaluation of the activation of cells.19-

22 The chemical structure of tetrazolium salt shows in Scheme 1-1. Tetrazolium salts are

quaternary ammonium compounds and soluble in water. Generally, three alkyl groups

contain aromatic rings to enhance hydrophobicity that allows to penetrate phospholipid

layer membrane.

Scheme 1-1. Chemical structure of the cationic tetrazolium salt and the reduced formazan

compound.

Mosmann firstly reported the colorimetric assay for assessing proliferation of

mammalian cell by using a yellow water-soluble tetrazolium dye, 3-(4,5-di-

methylthiazol-2-yl)-2,5-diphenyzl tetrazolium bromide (MTT).23 MTT is transformed

3

into to insoluble purple formazan compounds in by reaction with intracellular reductase

such as coenzyme nicotinamide adenine dinucleotide phosphate, and/or quinone species.

The colored precipitates are dissolved into organic solvent or oils (dimethyl sulfoxide,

dioxane or cyclohexane etc.) for evaluating the absorbance by spectroscopy and the

colored signal depend on the degree of activation of cells. MTT assay can detect only

living cell, and therefore quantification of cellular cytotoxicity or proliferation can be

measured. In addition, water-soluble formazan compounds with sulfone group introduced

have been developed to simplify the experimental procedure.24-26 Tsukatani et al reported

the colorimetric evaluation of bacterial growth by using a water-soluble

tetrazolium/formazan system.27 They indicated that a precursor of a water-soluble

formazan, 2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-

tetrazolium (WST-8), was superior to conventional tetrazolium salts including MTT with

the regard to the reactive efficiency with the intracellular reductase. However, long

incubation time of at least 4 h or more was required to obtain a sufficient colorimetric

signal at a lower cell density of 1.93×105 CFU mL－1.

Crystal violet (CV) is one of the triarylmethane dye widely used as a histological stain,

and shows blue color when dissolved in water. The most commonly used application of

CV dye is to distinguish bacterial species into two large group, Gram-positive and Gram-

negative.28,29 Gram-positive bacteria have a thick peptidoglycan layer in the cell wall that

retains CV staining, while Gram-negative bacteria have a thinner peptidoglycan layer that

allows to rinse remaining dye. Gram-negative bacteria, such as Escherichia coli and

Pseudomonas aeruginosa, cause infection disease and food poisoning due to toxic

lipopolysaccharide on the outer cell membrane. Therefore, CV stain is one of effective

methods as a first classification of determining whether bacteria has pathogenicity.

However, CV assay is insensitive to change in cell metabolic activity, and the color

change depend on the pH in the solution.30 Therefore, this technique may be inappropriate

for studies to quantitative assessment of cell metabolism.

1.2.2. Fluorometric assay

Fluorometric assays of cell activity are easily performed by using fluorescence

4

microscopy, fluorometer or flow cytometer. Fluorescence bioimaging is becoming one of

the most indispensable techniques that can visualize the dynamics of intracellular

molecule and cell activity in living tissues and organs.31-33 These assays more sensitive

than colorimetric assays, and various kits are commercially available for quantification

of specific biomolecule activity, including, glucose oxidase, lactate dehydrogenase,

nicotinamide adenine dinucleotide, and deoxyribonucleic acid.

Scheme 1-2. Chemical structure of the resazurin and the reduced resorufin compound.

Resazurin is a redox dye and metabolic indicator for quantitative estimation of viable

cells (Scheme 1-2).34,35 The protocol is similar to that utilizing the tetrazolium/formazan

system. The oxidized-form resazurin penetrated into intracellular environment, where it

is enzymatically reduced to fluorescent resorufin. The fluorescent signal is correlated with

viable cell number. However, the estimation is sometimes suspectable because of the

possibility of fluorescent interface from cellular compound, and toxic effects of resorufin

itself on the cells.36 In addition, long incubation times for several hours is required.

1.2.3. Luminometric assay

Luminometric assays provide a simple and real-time detection of cell activation.

Especially, adenosine triphosphate (ATP) test is one of the most famous luminometric

assay for rapidly quantification of cell activation. ATP is a biomolecule that exist in all

microorganism, fungi, bacteria, and mold to provide energy for metabolic processes such

as cell growth, respiration, fermentation, or photosynthesis. McElroy firstly reported that

ATP specifically react with firefly luciferin and appear a strong luminometric light which

intensity is proportional to the ATP concentration.37 The reaction does not require an

incubation step to produce colored compounds such as formazan or resorufin. Therefore,

ATP bioluminescence test make it possible to evaluate ATP concentration simply and

5

rapidly just to swipe environmental surface. ATP concentration is a critical indicator for

assessing microbial contamination. Therefore, ATP test has been applied various field

including food/beverage factory and healthcare facilitate. However, the luminometric

intensity varies greatly depending on the bacterial species, and poor detection of Gram-

negative bacteria.38,39 In addition, chloride ion in fungicides or disinfectants has as a

strong inhibitory effect on luminometric signals.40,41

1.3. Electrochemical techniques in bioanalysis

Conventional assay for evaluating metabolic activity, including colony-counting and

color-based technique, are complicated assays that consist of incubation,

centrifugation/filtration, and/or immobilization steps, which hamper obtaining precise

kinetic information about metabolic activity. Electrochemical techniques are reasonable

for monitoring cell activity because most of metabolic functions are composed of a series

of redox reactions. Real-time measurements of electrochemical signals make it possible

to quantify various biomolecules functions such as membrane protein or cell secretion in

intracellular/extracellular environment.

1.3.1. Voltammetric technique

Voltammetry is a basic electrochemical technique for measuring current through

electrode by sweeping potential, and one of the well-studied electrochemical tools for

investigations of the biological functions including microbe and enzyme.42-44 Cell-

immobilized electrode is often used as a working electrode for directly monitoring redox

reaction in intracellular/extracellular environment, and make it possible to quantify redox

biomolecules and electron transfer rates. Previous studies show the voltammetric analysis

for quantitative profiling of intracellular quinones.45,46 Quinone species in the

cytoplasmic membrane play important roles in microbial respiratory chain, and therefore

understanding of its dynamic behavior requires frequent monitoring. It was found that

intracellular quinone species are easily immobilized on an electrode by heating bacterial

suspension on the electrode. Cyclic voltammogram shows that two pairs of redox peaks,

6

each assigned to the adsorption of isoprenoid ubiquinone and menaquinone, giving

midpeak potentials of -0.015 and -0.25 V (vs. Ag|AgCl). It is noteworthy that total

quinone concentration in a single cell was constant during incubation, indicating that the

cytoplasmic membrane was saturated with quinone species. Furthermore, quinone

concentration in a single cell was estimated to be approximately 129 zmol cell－1. This

technique makes it possible to measure just 1 min by heat evaporation of a suspension

containing the targeted bacteria on the electrode.

Recently, metal-reducing bacteria receive considerable attention for various

utilization in bioelectrochemical and bioremediation processes. Shewanella oneidensis is

one of the most useful metal-reducing bacteria, and attracts broad attention because of its

unique metabolic capacities with regard to extracellular electron transfer to metal-ions

and/or electrode. Cytochrome c proteins are arranged on the inner/outer membrane, which

make it possible to directly transfer electrons to extracellular electron acceptor.47-49 On

the other hand, it was also reported that flavin species in cell secretion and biofilm

mediate electron transfer between bacteria and electron acceptor.50 Baron et al reported

voltammetric analysis to investigate the role of cytochrome c proteins in extracellular

electron transfer from Shewanella cell to electrode.51 S. oneidensis cells were

immobilized on the electrode with biofilm formation (1.0–3.0 μm thick layer). In the

absence of soluble flavins, electron transfer occurred in a broad potential window

centered at approximately 0 V (vs. Ag|AgCl). In contrast, the addition of soluble flavin

allowed to accelerate electron transfer to electrode at lower applied potentials around −0.2

V. These results suggested that anodic current in the higher >0 V window is due to

activation of a direct transfer mechanism by cytochrome c proteins, whereas electron

transfer at lower potentials is attributed to mediated electron transfer by soluble flavins

species. These studies indicated that voltammetric techniques are suitable to not only

quantify redox species in intracellular/extracellular environment but elucidate the

mechanism of electron transfer in bacterial cells.

7

1.3.2. Electrochemical microsystem technique

Minimamization of electrochemical measurement system make it possible to evaluate

metabolic activity at a single cell level. Amperometric techniques with micro/nano

electrode allows us to noninvasively probe redox species around a single cell.52,53

Matsumae et al performed real-time monitoring of intracellular quinone oxidoreductase

activity (NQO) of a living mammalian cell with cell-penetration mediator, menadione

that allows to mediate electron transfer between intracellular NQO and extracellular

electrode reaction.54 The amperometric observation shows that intracellular NQO activity

at a single cell was estimated to 7.15±3.54 × 10−17 mol s－1. Liu et al measured current

responses of directly electron transfer by S. oneidensis using optical tweezer, which

enable to contact physically of a single bacterial cell onto the microelectrode.55 It was

found that a single bacterial cell produced a current of approximately 200 fA at applying

potential of + 0.2 V (vs. Ag|AgCl) to electrode. These studies suggested that

electrochemical microsystem techniques are promising methods for evaluating cell

activity with high sensitivity.

1.4. Outline of the thesis

As mentioned above, electrochemical techniques are suitable methods for

quantification of cell activity with high sensitivity and in real-time. Therefore,

electrochemical analysis was performed for quantitative assessment of bacterial activity,

focusing on the redox species, protein, and electron transfer in intracellular/extracellular

environment.

Chapter 1 is introduction, which described about bacteria, color-based analysis of

metabolic activity, and electrochemical techniques in bioanalysis.

Chapter 2 shows a usefulness of microbial platform constructed by conductive

polymers for observation of bacterial activity. Conducting polymers works as a

biocompatible matrix for entrapping bacterial cells on an indium-tin-oxide (ITO)-coated

electrode. The cell density and viability were optically evaluated by microscopy. The

8

conducting polymers also facilitated electrochemical evaluation of the respiratory activity

of bacterial cells.

Chapter 3 demonstrates a viable bacterial detection focused on the electrochemical

property of tetrazolium salts, which was converted to an insoluble and redox active

formazan compound in viable microbial cells. I focused on not colorimetric but

electrochemical property of MTT. The insolubility of this formazan was effectively

exploited as a surface-confined redox event. Bacterial suspensions that incubated with a

tetrazolium salt was applied to ITO electrode and heat to dry for the adsorption. A

distinctive voltammetric oxidation peak appeared, and the magnitude was correlated to

the number of viable microbes in the suspension.

Chapter 4 describes the precious metal-ion reduction by S. oneidensis. I tracked the

formation of gold nanoparticles (Au NPs) on the S. oneidensis cell surfaces, and

investigated the roles of membrane proteins and extracellular polysaccharides in this

process. In addition, I propose a simple method for the detection of metal ions in solution,

focusing on the light-scattering characteristics of the metal nanoparticles formed on the

cells.

Chapter 5 denotes a quantitative evaluation of electron generation based on individual

enzyme reactions in S. oneidensis. By using potentiometric measurements, I have

examined intracellular electron generation in bacterial suspensions of S. oneidensis

supplemented with different carbon sources or ferricyanide. Real-time measurements by

potentiometry in bacterial suspensions enabled precise quantification of the number of

electrons generated by S. oneidensis based on the Nernst equation, because the

[ferricyanide]/[ferrocyanide] ratio immediately changed during the incubation.

Chapter 6 summarized the whole results and conclusions of the thesis.

9

REFERENCES

(1) Whitman, W. B.; Coleman, D. C.; Wiebe, W. J. Proc. Natl. Acad. Sci. U. S. A. 1998,

95, 6578.

(2) Kato, C.; Li, L.; Nogi, Y.; Nakamura, Y.; Tamaoka, J.; Horikoshi, K. Appl. Environ.

Microbiol. 1998, 64, 1510.

(3) Santillana, M. M.; Brito, E. M. S.; Duran, R.; Corona, F. G. Exermophiles, 2017, 21,

499.

(4) Segawa, T.; Ushida, K.; Narita, H.; Kanda, H.; Kohshimae, S. Polar Sci. 2010, 4, 215.

(5) Yoshida, S. Toda's New biacteriology, 34th ed.; Nanzando Co. Ltd.: Tokyo, 2013.

(6) Madigan, M. T.; Martinko, J. M.; Parker, J. Brock Biology of Microorganisms, 9th

ed.; Prentice Hall, INC., 2000.

(7) Sobsey, M. D. In Identifying Future Drinking Water Contaminants, Macdonald, J. A.;

Gibson, M.; Swartz, K. A. Eds.; National Academy Press: Washington, D.C., 1999.

(8) Kinoshita, T.; Nguyen, D. Q.; Le, D. Q.; Ishiki, K.; Shiigi, H.; Nagaoka, T. Anal.

Chem. 2017, 89, 4680.

(9) Wildt, R. M. T.; Mundy, C. R.; Gorick, B. D.; Tomlinson I. M. Nat. Biotechnol. 2000,

18, 989.

(10) Ding, T.; Bilitewski, U.; Schmid, R. D.; Korz, D. J.; Sanders, E. A. J. Biotechnol.

1993, 27, 143.

(11) Kau, A. L.; Ahern, P. P.; Griffin, N. W.; Goodman, A. L.; Gordon, J. I. Nature, 2011,

474, 327.

(12) Montagne, L.; Pluske, J. R.; Hampson D. J. Anim. Feed Sci. Technol. 2003, 108, 95.

(13) Kim, H. J.; Park, H. S.; Hyun, M. S.; Chang, I. S.; Kim, M.; Byung, K. H. Enzyme

Microb. Technol. 2002, 30, 145.

10

(14) Konishi, Y.; Tsukiyama, T.; Tachimi, T.; Saitoh, N.; Nomura, T.; Nagamine, S.

Electrochim. Acta 2007, 53, 186.

(15) Xu, J.; He, W.; Wang, Z.; Zhang, D.; Sun, J.; Zhou, J.; Li, Y.; Su, X. Front. Bioeng.

Biotechnol. 2016, 4, 86.

(16) Simpson, A. J.; Maxwell, A. I.; Govan, J. R. W.; Haslett, C.; Sallenave, J. M. FEBS

lett. 1999, 452, 309.

(17) Hattori, H. Soil Sci. Plant Nutr. 1992, 38, 93.

(18) Thomas, S.; Andrews, A. M.; Hay, N. P.; Bourgoise, S. J. Tissue Viability 1999, 9,

127.

(19) Stubberfield, L. C. F.; Shaw, P. J. A. J. Microbiol. Methods 1990, 12, 151.

(20) Hatzinger, P. B.; Palmer, P.; Smith, R. L.; Peñarrieta, C. T.; Yoshinari, T. J. Microbiol.

Methods 2003, 52, 47.

(21) Roehm, N. W.; Rodgers, G. H.; Hatfield, S. M.; Glasebrook, A. L. J. Immunol.

Methods 1991, 142, 257.

(22) Posch, T; Pernthaler, J.; Alfreider, A.; Psenner, R. Appl. Environ. Microbiol. 1997,

63, 867.

(23) Mosmann, T. J. Immunol. Methods, 1983, 65, 55.

(24) Roslev, P.; King, G. M. Appl. Environ. Microbiol. 1993, 59, 2891.

(25) Tominaga, H.; Ishiyama, M.; Ohseto, F.; Sasamoto, K.; Hamamoto, T.; Suzuki, K.;

Watanabe, M. Anal. Commun. 1999, 36, 47.

(26) Buttke, T. M.; McCubrey, J. A.; Owen, T. C. J. Immunol. Methods 1993, 157, 233.

(27) Tsukatani, T.; Higuchi, T.; Suenaga, H.; Akao, T.; Ishiyama, M.; Ezoe, T.; Matsumoto,

K. Anal. Biochem. 2009, 393, 117.

(28) Fischer, R.; Larose, P. J. Bacteriol. 1952, 64, 435.

11

(29) O'Toole, G. A. J. Bacteriol. 2016, 198, 3128.

(30) Sarma, G. K.; Gupta, S. S.; Bhattacharyyaa, K. G. J. Environ. Manage. 2016, 171,

1.

(31) Inada, N.; Fukuda, N.; Hayashi, T.; Uchiyama. S. Nat. Protoc. 2019, 14, 1293.

(32) Hu, F.; Huang, Y.; Zhang, G.; Zhao, R.: Yang, H.; Zhang, D. Anal. Chem. 2014, 86,

7987.

(33) Chan, J.; Dodani, S. C.; Chang, C. J. Nat. Chem. 2012, 4, 973.

(34) Mariscal, A.; Lopez-Gigosos, R. M.; Carnero-Varo, M.; Fernandez-Crehuet, J. Appl.

Microbiol. Bitechnol. 2009, 82, 773.

(35) Xiao, J.; Zhang, Y.; Wang, J.; Yu, W.; Wang, W.; Ma, X. Appl. Biochem. Biotechnol.

2012, 162, 1996.

(36) Pace, R. T.; Burg, K. J. L. Cytotechnol. 2015, 67, 13.

(37) McElroy, W. D. Proc. Natl. Acad. Sci. USA, 1947, 33, 342.

(38) Selan, L.; Berlutti, F.; Passariello, C.; Thaller, M. C.; Renzini, G. J. Clin. Microbiol.

1992, 30, 1739.

(39) Turner, D. E.; Daugherity, E. K.; Altier, C.; Maurer, K. J. J. Am. Assoc. Lab. Anim.

2010, 49, 190.

(40) Strehler, B. L.; Totter, J. R. Arch. Biochem. Biophys. 1952, 40, 28.

(41) Omidbakhsh, N.; Ahmadpour, F.; Kenny, N. PloS One. 2014, 9, e99951.

(42) Ikeda, T.; Kurosaki, T.; Takamasa, K.; Kano, K. Anal. Chem. 1996, 68, 192.

(43) Gulaboski, R.; Mirčeski, V.; Bogeski, I.; Hoth, M. J. Solid State Elecrochem. 2012,

16, 2315.

(44) Xiao, Y.; Zhang, E.; Zhang, J.; Dai, Y.; Yang, Z.; Christensen, H. E. M.; Ulstrup, J.;

Zhao, F. Sci. Adv. 2017, 3, e1700623.

12

(45) Morishita, A.; Higashimae, S.; Nomoto, A.; Shiigi, H.; Nagaoka, T. J. Electrochem.

Soc. 2016, 163, G166.

(46) Le, D. Q.; Morishita, A.; Tokonami, S.; Nishino, T.; Shiigi, H.; Miyake, M.; Nagaoka,

T. Anal. Chem. 2015, 87, 8416.

(47) Fredrickson, J. K.; Romine, M. F.; Beliaev, A. S.; Auchtung, J. M.; Driscoll, M. E.;

Gardner, T. S.; Nealson, K. H.; Osterman, A. L.; Pinchuk, G.; Reed, J. L.; Rodionov, D.

A.; Rodrigues, J. L.; Saffarini, M. D. A.; Serres, M. H.; Spormann, A. M.; Zhulin, I. B.;

Tiedje, J. M. Nat. Rev. Microbiol. 2008, 6, 592.

(48) Okamoto, A.; Nakamura, R.; Hashimoto, K. Electrochim. Acta, 2011, 56, 5526.

(49) Carmona-Martinez, A. A.; Harnisch, F.; Fitzgerald, L. A.; Biffinger, J. C.; Ringeisen,

B. R.; Schröder, U. Bioelectrochem. 2011, 81, 74.

(50) Marsili, E.; Baron, D. B.; Shikhare, I. D.; Coursolle, D.; Gralnick, J. A.; Bond, D. R.

Proc. Natl. Acad. Sci. U. S. A. 2008, 1055, 3968.

(51) Baron, D.; LaBelle, E.; Coursolle, D.; Gralnick, J. A.; Bond, D. R. J. Biol. Chem.

2009, 284, 28865.

(52) Actis, P.; Tokar, S.; Clausmeyer, J.; Babakinejad, B.; Mikhaleva, S.; Cornut, R.;

Takahashi, Y.; Córdoba, A. L.; Novak, P.; Shevchuck, A. I.; Dougan, J. A.; Kazarian, S.

G.; Gorelkin, P. V.; Erofeev, A. S.; Yaminsky, I. V.; Unwin, P. R.; Schuhmann, W.;

Klenerman, D.; Rusakov, D. A.; Sviderskaya, E. V.; Korchev, Y. E. ACS Nano 2014, 8,

875.

(53) Huang, Y.; Cai, D.; Chem, P. Anal. Chem. 2011, 83, 4393.

(54) Matsue, Y.; Takahashi, Y.; Ino, K.; Shiku, H.; Matsue, T.; Anal. Chim. Acta 2014,

842, 20.

(55) Liu, H.; Newton, G. J.; Nakamura, R.; Hashimoto, K.; Nakanishi, S.; Angew. Chem.

Int. Ed. 2010, 49, 6596.

13

Chapter II

A Microbial Platform Based on Conducting Polymers for Evaluating Metabolic Activity

2.1. Introduction

Although pathogenic bacteria pose a threat to human life by causing food poisoning1

and infectious diseases,2,3 some strains find application in areas such as sewage

purification,4,5 decomposition of toxic substances,6,7 and microbial fuel cell

construction.8−10 Hence, a better understanding of the biological functions of bacteria is

required to reduce the associated threat and increase the usefulness of these

microorganisms, which, in turn, necessitates the quantitative evaluation of bacterial

metabolic processes including growth and respiration.

Bacterial growth and respiration have been assessed by cell-counting, staining, or gas

chromatography.11−14 Conventional methods require one to individually prepare

substrates and containers such as agar plates, flasks, and microplates for bacterial samples

according to the purpose and method. One means of assessing bacterial metabolism is by

immobilizing these entities on a glass slide or an electrode. Immobilization allows

bacterial properties to be monitored by optical and/or electrochemical methods and can

serve as a platform for carrying out biocatalytic metabolite production and constructing

biofuel cells and biosensors.15 In many cases, living cells must be stabilized by

confinement in a suitable matrix. The long-term viability and metabolic activity of

confined bacteria are influenced by the chemical and mechanical properties of the matrix,

including biocompatibility, structural porosity, and contraction. Immobilization of live

bacteria is essential for reliable evaluation of their activity through microscopic

observations and electrochemical measurements. From such a point of view, the

usefulness of the conductive polymer as a matrix material has been clarified. The

effectiveness of conducting PPy-modified gold substrates in adhesion and proliferation

of mouse stem cells has been demonstrated.16 PPy also indicated a good biocompatibility

for various microorganisms such as fungi,17,18 yeast,19−21 and bacterial cells22 in the

14

improvement of mechanical properties of cells and charge transfer efficiency from cells.

Previous studies indicated that conducting polymers allowed to immobilize bacterial

cells on substrates.23−27 Bacterial cells are retained by the polymer matrix as anionic

dopants due to the negative zeta potential generated by the phosphate groups on the

lipopolysaccharides that comprise the outer membrane of the cell. Herein, I demonstrate

the utility of conducting polymer films as a matrix for evaluating the biological properties

by monitoring the growth and respiration of cells immobilized on conducting polymer-

coated ITO substrates.

2.2. Experimental

2.2.1. Chemicals and materials

All chemical reagents were of analytical grade and were used as supplied without

further purification, unless indicated otherwise. Ultrapure water (resistance >18 MΩ) was

used in all experiments. Pyrrole was purchased from Wako Pure Chemical Industries

(Japan), and 3,4-ethylenedioxythiophene (EDOT) was obtained from Sigma-Aldrich.

Nutrient broth (NB) was obtained from Eiken Chemicals (E-MC35, Japan). Escherichia

coli, Acetobacter xylinum, Pseudomonas aeruginosa, and Salmonella enterica were

acquired from the Biological Resource Center, National Institute of Technology and

Evaluation (NBRC, NITE, Japan).

2.2.2. Bacterial culturing

All experiments involving bacterial cultures were executed in a biosafety level 2

laboratory and designed and managed in accordance with safety regulations. A. xylinum

was incubated in a medium agar plate (NBRC 350 medium) at 30 °C for 3 d. A single

colony on the plate was suspended in 30 mL of liquid 350 medium. After cultivation, the

suspension of A. xylinum (7.5 mL) was added to a glass flask containing 142.5 mL of the

liquid 350 medium and incubated for 2 d. Other bacteria were incubated in an NB agar

plate at 30 °C for 24 h. A single colony on the plate was suspended in 30 mL of liquid

NB medium and cultured overnight upon shaking. The bacterial suspension then was

15

centrifuged at 8000 rpm (7510g) for 5 min, the supernatant was removed, and the

precipitate was resuspended in sterilized water. The procedure was repeated twice to

obtain purified bacterial suspensions (1.0 × 109 CFU mL–1).

2.2.3. PEDOT film fabrication

The surface of an ITO glass strip with dimensions of 26 × 77 mm2 (Kinoene Optics

Inc., Japan) was covered with a UV-curing resin film using a Roland DG LEF12 inkjet

printer, with the exception of a circular area (0.79 cm2) used as the working electrode.19−22

A platinum mesh and a Ag|AgCl electrode (filled with 3.0 M KCl) were used as counter

and reference electrodes, respectively. The electrodes were placed in a glass cell that

contained a solution (6.0 mL) of EDOT (10 mM) and purified A. xylinum in a phosphate

buffer (pH 5.3). Electrochemical polymerization was carried out at +1.05 V (vs Ag|AgCl)

for 900 s to obtain a cell-doped PEDOT film on the ITO electrode. The freshly prepared

A. xylinum/PEDOT film was then rinsed with a copious amount of sterile water.

2.2.4. PPy film fabrication

Pyrrole (100 mM) was added to the as-prepared bacterial suspension of E. coli, P.

aeruginosa, or S. enterica in a phosphate buffer (pH 3.0). An ITO electrode covered with

an insulating film to control the electrode area (0.13 cm2) was used as the working

electrode.26−29 Electrochemical polymerization was carried out at +0.98 V (vs Ag|AgCl)

for 100 s in the phosphate buffer containing bacterial cells to obtain a cell-doped PPy film.

The PPy film-coated ITO electrode was rinsed with sterile water and then used as the

working electrode.

2.2.5. Spectroscopy

The A. xylinum/PEDOT-modified ITO glass strip was subjected to UV–vis

spectrometry (V-730, Jasco, Japan) to confirm the formation of PEDOT. The glass strip

was immersed into 30 mL of liquid 350 medium and cultured for 1 d. During incubation,

the strip was removed from the medium, set in the sample chamber, and subjected to

absorption spectrum measurement. Identical operations were carried out for a blank strip

on which PEDOT was formed without A. xylinum.

16

2.2.6. Microscopic observations

The bacteria-doped conducting polymer film was observed using a dark-field

microscope (ECLIPSE Ni, Nikon, Japan) equipped with a dark-field condenser, a 100-W

halogen lamp, and a camera with a charge-coupled device (DS-Ri1, Nikon, Japan).30,31

Bacterial viability was evaluated by counting stained cells with a fluorescent microscope

according to the manufacturer’s instructions for the BacLight bacterial viability kit

(ThermoFisher Scientific). The above kit contained two fluorescent pigments, SYTO9,

which strained both living and dead cells, and propidium iodide, which strained only dead

cells.

2.2.7. Electrochemical measurements

Scheme 2-1. A Custom-Made Thin-Layer Electrolytic Cella

aInset shows a photograph of the PPy-modified ITO glass.

Voltammetric measurements were performed using a custom-made thin-layer

electrolytic cell (Scheme 2-1). A piece of filter paper (No. 1, φ 55 mm, Toyo Roshi Kaisha,

Ltd., Japan) folded in half was placed on the PPy-modified ITO electrode. An appropriate

amount of phosphate buffer (0.15 mL) was dropped on the filter paper, and another ITO

glass slide was placed on the filter paper as a counter electrode. The assembly was fixed

with an adhesive Teflon tape, and a Ag|AgCl reference electrode was inserted into the

folded filter paper. Cyclic voltammograms (CVs) were recorded between −0.8 and +0.6

V at a scan rate of 10 mV s–1 under aerobic conditions at 310 K. The concentration of

dissolved oxygen in the buffer was measured using an oxygen sensor (Firesting O2,

PyroScience GmbH, Germany).

17

2.3. Results and Discussion

Figure 2-1. (A) SEM and (B) dark-field images of an A. xylinum/PEDOT film incubated

in liquid 350 medium for (a) 0, (b) 12, (c) 24, (d) 42, and (e) 48 h under aerobic conditions.

(f) SEM image of the cellulose layer peeled from the PEDOT film. Scale bars are 10 μm.

The insets show photographs of the PPy-modified ITO glass and the cellulose layer. (C)

Time dependence of the cell density of A. xylinum (n = 3). (D) UV–vis spectra of PEDOT

(a) with and (b) without A. xylinum incubated in liquid 350 medium for 24 h under aerobic

conditions.

18

The scanning electron microscopy (SEM) image in Figure 2-1Aa shows the A.

xylinum cells incorporated in a PEDOT film. The difference in the contrast of the image

arises from the difference in the electrical conductivity of the film components. The

conductive PEDOT matrix appears light gray, whereas the insulating bacterial cells are

dark gray. Most bacterial cells are rod-shaped, and their average width and length are 1.5

and 4.0 μm, respectively. The estimated population density is 8.8 × 105 cells cm–2. A

dark-field image clearly demonstrates the presence of A. xylinum cells in the film (Figure

2-1Ba). The cells consist of 70% water and appear bright in rod-like form due to their

greater incident-light scattering ability based on the difference in the refractive index

between PEDOT (>1.5) and water (1.3) in the cell.32

Figure 2-2. Fluorescent images of the as prepared A. xylinus-doped PEDOT film stained

using the LIVE/DEAD BacLight Bacterial Viability Kit, (a) excitation with blue light

(465–495 nm) and (b) green light (525-540 nm). Living cells stained with Syto9 were

observed as green spots and dead cells were observed as red spots with propidium iodide.

Scale bars are 50 μm.

The estimated viability of A. xylinum cells immobilized in PEDOT is greater than

90% (Figure 2-2). Although there is good correlation between the SEM and dark-field

images of the film, the high vacuum condition results in a significant decrease in bacterial

viability (≃0%). The ability to observe the bacteria at normal temperature and pressure

using dark-field microscopy makes it possible to follow the cell growth without affecting

its viability.

19

Further, an A. xylinum/PEDOT film-coated ITO glass was immersed in liquid 350

medium and incubated aerobically at 303 K. After incubation, the film was rinsed with

sterilized water and placed on the stage of a dark-field microscope. The cell density

increased to 2.4 × 106 cells cm–2 after 12 h (Figure 2-1Bb). A. xylinum cells synthesize

cellulose nanofibers from glucose.33 The production of bacterial cellulose known as a

component of biofilm is very evident in the SEM image (Figure 2-1Ab). Although a cell

density of 1.6 × 107 cells cm–2 was observed in the dark-field image after 24 h (Figure 2-

1C), it became difficult to estimate the cell number based on SEM imaging due to the

extensive production of cellulose nanofibers (Figure 2-1Ac) that ultimately covered the

cells. After 42 h, the continued production of cellulose nanofibers leads to the formation

of a white film (inset, Figure 2-1Ad), which prevents the observation of cells by dark-

field microscopy (Figure 2-1Bd). Peeling the cellulose layer from the PEDOT film after

48 h incubation leads to a cell density of 2.7 × 106 cells cm–2 in the film adhered to the

ITO glass. This value is comparable to the cell density in the original film (Figure 2-1Be).

The SEM image of the cellulose layer reveals many A. xylinum cells (Figure 2-1Af). Cell

growth occurs in the cellulose layer, but not in the liquid 350 medium. These observations

are consistent with the formation of a biological film of cellulose fibers produced by A.

xylinum cells. The UV–vis spectrum of an A. xylinum/PEDOT film obtained before

incubation (0 h) exhibited a broad absorption at 700 nm well matched with that of the

PEDOT film without A. xylinum (Figure 2-1D). Thus, bacteria did not affect the chemical

structure of PEDOT. The absorption at 700 nm decreases in intensity and shifts toward

500 nm during 24 h incubation. This behavior suggests that the bipolaron population

decreases and the polaron population increases in the polymer backbone, thus initiating a

change in PEDOT from a conducting to insulating state. Thus, bacterial growth and

biofilm formation do not affect the chemical structure of PEDOT, as a PEDOT film

without bacteria exhibits a similar spectral change.

20

Figure 2-3. (A) SEM images of an E. coli/PPy film: (a) top view and (b) 89°-angle view.

(B) Dark-field images of a PPy film incubated aerobically in liquid NB medium for (a) 0,

(b) 9, and (c) 18 h. (C) Time-dependence of the cell density of E. coli under aerobic and

anaerobic conditions in (a) a PPy film excluding biofilm regions and (b) liquid NB

medium (n = 3).

21

The SEM image of a conducting PPy film on ITO reveals the presence of 1.2 × 2.5

μm E. coli cells (Figure 2-3A) roughly half-embedded in the ∼800 nm-thick PPy film.34

The dark-field image shows the cells as light, rod-shaped bodies due to the difference in

the refractive index between water and PPy (>1.5).35 The cells are well dispersed (Figure

2-3B) and have a density of 1.5 × 106 cells cm–2. Fluorescence observation indicates the

cell viability to be greater than 99%. Therefore, PPy provides a favorable environment

for bacteria, due to its greater biocompatibility than PEDOT.22 Growth of the facultative

anaerobe, E. coli, was examined by dark-field microscopy. An E. coli/PPy film was

immersed in liquid NB medium and incubated aerobically at 303 K. The cell density

increased slightly to 1.6 × 106 cells cm–2 after 6-h incubation. Bacterial growth occurs in

four stages comprising (i) lag, (ii) log, (iii) stationary, and (iv) death phases. No bacterial

growth occurs during the lag phase, because essential cellular components including RNA

and enzymes must first be synthesized.36 A significant increase in the cell density was

observed after 9 h, indicating that E. coli had entered the log phase with cell division

(Figure 2-3Bb). Biofilm formation was evident at places where bacteria had gathered.

The estimated cell density in regions apart from the biofilm is 5.3 × 106 cells cm–2 (Figure

2-3 C). The cell density ultimately reached 3.2 × 107 cells cm–2 excluding the biofilm

region after 18 h with a marked increase in the amount of biofilm by this time. It is clear

that it shows a more rapid growth curve, considering the number of bacteria in the biofilm

during the period of 9–18 h. E. coli cell growth entered the stationary phase between 18

and 24 h (3.7 × 107 cells cm–2). I presume that the growth in the stationary phase is limited

by the depletion of an essential nutrient.37,38 The cell viability gradually decreased

thereafter, and almost all bacterial cells were dead after 48 h.

This typical growth pattern suggests that E. coli cells maintain metabolic activity in

the PPy matrix as well as in the liquid medium (Figure 2-3Cb). The result establishes PPy

as a suitable environment for bacterial growth and confirms its combination with dark-

field microscopy as a powerful tool for observing live cells under normal atmospheric

conditions. E. coli cells in PPy films exhibit similar behavior under anaerobic conditions,

in line with the fact that these bacteria are facultative anaerobes that can sustain

22

metabolism under both aerobic and anaerobic conditions. The estimated cell density of

8.0 × 106 cells cm–2 after 18-h incubation is ∼4 times less than that obtained under aerobic

conditions (Figure 2-3Ca). Bacterial growth is strongly correlated with the level of

adenosine triphosphate (ATP) as an energy source.39 In turn, ATP generation within a cell

depends on the concentration of oxygen, which, as the final electron acceptor, drives the

tricarboxylic acid cycle and accounts for greater ATP production under aerobic

conditions (36 mol) than under anaerobic ones (2.0 mol) conditions.40,41 Therefore, E. coli

cells grow more rapidly in the presence of oxygen.

Figure 2-4. (A) Dark-field images of P. aeruginosa in a PPy film. The PPy film incubated

aerobically in liquid NB medium for (a) 0, (b) 9, and (c) 18 h at 303 K. Scale bars are 200

μm. (B) Time-dependence of the cell density of P. aeruginosa in (a) PPy film (left axis)

and (b) liquid NB medium (right axis). Number of experiments was 3.

P. aeruginosa is an aerobic bacterium that exhibits metabolic activity in the presence

of oxygen. Its cell density in a PPy film (Figure 2-4A) increases from an initially small

value (1.5 × 106 cells cm–2) to 2.6 × 107 cells cm–2 after 9-h incubation (Figure 2-4Ba).

The cell density in the film decreases after 12 h, but the number of bacteria in the liquid

medium increases significantly after this time (Figure 2-4Bb). This indicates that

proliferated P. aeruginosa cells migrated from the PPy film to the liquid medium because

23

P. aeruginosa growth is better sustained in the aerobic liquid medium than in the

relatively anaerobic atmosphere of the PPy film.

Figure 2-5. (A) Dark-field images of a PPy film prepared by electrochemical

polymerization for (a) 100 and (b) 300 s. Scale bars are 10 μm. The inset shows a cross-

sectional SEM image of the film with marked thickness. (B) Dependence of E. coli cell

density on the polymerization time used to obtain the PPy film (n = 5). (C) Aerobic CVs

of a live E. coli/PPy film recorded in the thin-layer electrolysis cell containing 20 mM

glucose (a) before incubation and (b,c) after 30 min incubation. Polymerization times of

(a,b) 100 s and (c) 300 s were used.

I evaluated the respiratory activity of bacterial cells in a PPy film using a thin-layer

electrochemical cell (Scheme 2-1). I controlled the polymerization time for the formation

of the PPy film on the ITO glass to regulate the cell density of E. coli. An increase in the

thickness of the PPy film increased the apparent cell density on the ITO electrode (Figure

2-5A and B).26 I prepared the E. coli/PPy film with a polymerization time of 100 s (1.5 ×

106 cells cm–2). In the initial CV recorded between +0.6 and −0.8 V vs Ag|AgCl, the

Faradaic current that flows at potentials more negative than −0.3 V results from the

reduction of dissolved oxygen in the electrolyte solution. There was no discernible

24

difference in the CVs recorded with and without glucose at 0 min incubation time (Figure

2-5Ca). However, the oxygen reduction current decreased dramatically in the presence of

glucose after 30 min (Figure 2-5Cb). This change in current was not observed for the E.

coli/PPy film in glucose-free electrolyte or for a PPy film without E. coli in a glucose-

containing electrolyte. These results establish that the E. coli cells in the PPy film

consume dissolved oxygen and utilize it to oxidize glucose according to the following

equation:

C6H12O6 + 6H2O + 6O2 → 6CO2 + 12H2O (1)

The decrease in the Faradaic current between −0.3 and −0.8 V is attributed to the

consumption of dioxygen by E. coli based on eq 1. The accumulated charge of 20 μC

estimated from the difference in the current responses obtained before and after

incubation for 30 min equals the pink-colored area shown in Figure 2-5C. The quantities

of electron and oxygen were calculated to be 0.21 and 0.11 nmol, respectively, based on

eq 2.

O2 + 2H+ + 2e－ → H2O2 (2)

The estimated oxygen consumption by E. coli cells is 1.8 × 10–17 mol min–1 per cell.

I found that the dissolved oxygen in the electrolyte solution decreased with an increase in

the cell density of E. coli, which strongly affected the electrochemical response, and the

oxygen was completely consumed after the incubation of a PPy film prepared by

electrochemical polymerization for 300 s (2.6 × 106 cells cm–2).

Oxygen consumption by suspended E. coli cells was measured to confirm the above

result. An E. coli suspension (4.2 × 105 cells in 8 mL) was placed in a sealed container

equipped with a fiber optic oxygen sensor. The concentration of the dissolved oxygen in

the suspension decreased gradually in the absence of glucose due to its reaction with

stored glycogen (Figure 2-6).42 Linear correlation between the concentration of dissolved

oxygen and incubation time yields an oxygen reduction rate of 4.1 × 10–12 mol min–1. The

oxygen reduction rate in the E. coli suspension including glucose (20 mM) is 1.2 × 10–11

mol min–1. Therefore, the net oxygen reduction rate for glucose utilization by E. coli is

25

8.0 × 10–12 mol min–1. The per cell rate of oxygen consumption by suspended E. coli is

1.9 × 10–17 mol min–1 per cell, which agrees with the electrochemical result. The same

electrochemical experiments performed with S. enterica and P. aeruginosa demonstrate

that facultative anaerobic and aerobic bacteria exhibit similar oxygen consumption

(Figure 2-7).

Figure 2-6. Dissolved oxygen concentration in a stirred bacterial suspension (8.0 mL)

with and without 20 mM glucose. (A) E. coli (4.2 × 105 cells), (B) S. enterica (5.4 × 105

cells), (C) P. aeruginosa (2.1× 105 cells).

Figure 2-7. Aerobic cyclic voltammograms of (A) E. coli (1.6 × 105 cells)-, (B) S.

enterica (1.3 × 105 cells)-, (C) P. aeruginosa (1.5 × 105 cells)-immobilized PPy films

recorded in the thin-layer electrolysis cell containing 20 mM glucose during incubation

for 30 min. Polymerization times of 100 s were used, respectively.

The oxygen reduction current decreased with time in the presence of glucose. They

exhibit similar respiratory activities under aerobic conditions and high reproducibility

with variations within 10% of viability (Table 2-1), and are in good agreement with the

results obtained by optical sensor (Table 2-2). I also found that there was a large

26

difference in the voltammograms between A. xylinum and the others. This is expected to

be due to differences in glucose metabolism and should be investigated in detail by adding

organics to replace glucose. These results establish PPy film as an appropriate platform

for bacterial immobilization and activity monitoring.

Table 2-1. Oxygen consumption per single cell obtained by electrochemical

measurements.

aNumber of cells immobilized on the PPy film. bThe difference in the electric charge was

obtained with and without the addition of glucose. cRespiratory activity was calculated

by dividing the moles of consumed oxygen by the cell number.

Bacteria Number of cellsa

/cells

Electric chargeb

/C

Respiratory activityc

/mol min−1 cell−1

E. coli 2.0 × 105 2.0 × 10−5 1.8 × 10−17

1.6 × 105 1.3 × 10−5 1.4 × 10−17

S. enterica 1.8 × 105 1.5 × 10−5 1.5 × 10−17

1.3 × 105 1.2 × 10−5 1.6 × 10−17

P. aeruginosa 2.0 × 105 1.5 × 10−5 1.3 × 10−17

1.5 × 105 1.3 × 10−5 1.5 × 10−17

27

Table 2-2. Oxygen consumption per single cell obtained by optical sensor

aNumber of cells in an eight mL of bacterial suspension. bAmount of consumed oxygen. cRespiratory activity was calculated by dividing the moles of consumed oxygen by the

cell number.

2.4. Conclusion

I successfully measured bacterial activities in conducting polymer films to show that

this material provides a suitable environment for evaluating biological processes

including bacterial growth and biofilm formation. Bacterial growth in the film equals that

in a liquid medium. The conducting polymer matrix is a useful platform for evaluating

the metabolic activity of facultative anaerobic bacteria. In addition, its biocompatibility

and electrical conductivity facilitates quantitative evaluation of oxygen consumed by

bacterial cells. These findings are applicable to the analysis of living cells and to the

development of electrode materials for biofuel cells and biosensors modified with

exoelectrogenic bacteria such as Shewanella or Geobacter species.

Bacteria Number of cellsa

/cells

Consumed oxygenb

/mg L−1 min−1

Respiratory activityc

/mol min−1 cell−1 S.D. / %

E. coli

2.4 × 105 1.9 × 10−5 2.0 × 10−17

12 4.2 × 105 3.2 × 10−5 1.9 × 10−17

6.9 × 105 6.6 × 10−5 2.4 × 10−17

S. enterica

6.7 × 105 6.7 × 10−5 2.5 × 10−17

17 6.0 × 105 5.5 × 10−5 2.3 × 10−17

5.4 × 105 3.8 × 10−5 1.8 × 10−17

P. aeruginosa

2.1 × 105 1.3 × 10−5 1.5 × 10−17

47 4.5 × 105 1.9 × 10−5 1.0 × 10−17

8.3 × 105 8.7 × 10−5 2.6 × 10−17

28

REFERENCES

(1) Loir, Y. L.; Baron, F.; Gautier, M. Genet. Mol. Res. 2003, 2, 63.

(2) Costerton, J. W.; Stewart, P. S.; Greenberg, E. P. Science 1999, 284, 1318.

(3) Clemente, J. C.; Ursell, L. K.; Parfrey, L. W.; Knight, R. Cell 2012, 148, 1258.

(4) Rudolfs, W.; Falk, L. L.; Ragotzkie, R. A. Sewage Ind. Wastes 1950, 22, 1261.

(5) Painter, H. A. J. Gen. Microbiol. 1954, 10, 177.

(6) Gupta, S.; Dikshit, A. K. J. Biopestic. 2010, 3, 186.

(7) Dutta, S. World J. Pharm. Pharm. Sci. 2015, 4, 250.

(8) Kim, H. J.; Park, H. S.; Hyun, M. S.; Chang, I. S.; Kim, M.; Kim, B. H. Enzyme

Microb. Technol. 2002, 30, 145.

(9) Rabaey, K.; Boon, N.; Siciliano, S. D.; Verhaege, M.; Verstraete, W. Appl. Environ.

Microbiol. 2004, 70, 537.

(10) Zhou, M.; Chi, M.; Luo, J.; He, H.; Jin, T. J. Power Sources 2011, 196, 4427.

(11) Hrynkiewicz, K.; Baum, C.; Leinweber, P. J. Plant Nutr. Soil Sci. 2010, 173, 747.

(12) Ishiguro, K.; Washio, J.; Sasaki, K.; Takahashi, N. J. Microbiol. Methods 2015, 115,

22.

(13) Yu, L.; Yan, X.; Ye, C.; Zhao, H.; Chen, X.; Hu, F.; Li, H. PLoS One 2015, 10,

e0134401.

(14) Lobritz, M. A.; Belenky, P.; Porter, C. B. M.; Gutierrez, A.; Yang, J. H.; Schwarz,

E. G.; Dwyer, D. J.; Khalil, A. S.; Collins, J. J. Proc. Natl. Acad. Sci. U. S. A. 2015, 112,

8173.

(15) Gisela, S. A.; Maria, L. F.; Guillermo, J. C.; Martin, F. D.; Luis, E. D. Appl.

Microbiol. Biotechnol. 2009, 82, 639.

29

(16) Vaitkuviene, A.; Ratautaite, V.; Mikoliunaite, L.; Kaseta, V.; Ramanauskaite, G.;

Biziuleviciene, G.; Ramanaviciene, A.; Ramanavicius, A. Colloids Surf. A 2014, 442, 152.

(17) Kisieliute, A.; Popov, A.; Apetrei, R.-M.; Cârâc, G.; Morkvenaite-Vilkonciene, I.;

Ramanaviciene, A.; Ramanavicius, A. Chem. Eng. J. 2019, 356, 1014.

(18) Apetrei, R.-M.; Carac, G.; Ramanaviciene, A.; Bahrim, G.; Tanase, C.;

Ramanavicius, A. Colloids Surf., B 2019, 175, 671.

(19) Apetrei, R.-M.; Carac, G.; Bahrima, G.; Ramanaviciene, A.; Ramanavicius, A.

Bioelectrochemistry 2018, 121, 46.

(20) Andriukonis, E.; Stirke, A.; Garbaras, A.; Mikoliunaite, L.; Ramanaviciene, A.;

Remeikis, V.; Thornton, B.; Ramanavicius, A. Colloids Surf. B 2018, 164, 224.

(21) Ramanavicius, A.; Andriukonis, E.; Stirke, A.; Mikoliunaite, L.; Balevicius, Z.;

Ramanaviciene, A. Enzyme Microb. Technol. 2016, 83, 40.

(22) Stirke, A.; Apetrei, R.-M.; Kirsnyte, M.; Dedelaite, L.; Bondarenka, V.; Jasulaitiene,

V.; Pucetaite, M.; Selskis, A.; Carac, G.; Bahrim, G.; Ramanavicius, A. Polymer 2016,

84, 99.

(23) Shan, X.; Yamauchi, T.; Shiigi, H.; Nagaoka, T. Anal. Sci. 2018, 34, 483.

(24) Shan, X.; Yamauchi, T.; Yamamoto, Y.; Shiigi, H.; Nagaoka, T. Analyst 2018, 143,

1568.

(25) Shan, X.; Yamauchi, T.; Yamamoto, Y.; Niyomdecha, S.; Ishiki, K.; Le, D. Q.; Shiigi,

H.; Nagaoka, T. Chem. Commun. 2017, 53, 3890.

(26) Le, D. Q.; Tokonami, S.; Nishino, T.; Shiigi, H.; Nagaoka, T. Bioelectrochemistry

2015, 105, 50.

(27) Le, D. Q.; Takai, M.; Suekuni, S.; Tokonami, S.; Nishino, T.; Shiigi, H.; Nagaoka,

T. Anal. Chem. 2015, 87, 4047.

(28) Le, Q. D.; Morishita, A.; Tokonami, S.; Nishino, T.; Shiigi, H.; Miyakae, M.;

30

Nagaoka, T. Anal. Chem. 2015, 87, 8416.

(29) Morishita, A.; Higashimae, S.; Nomoto, A.; Shiigi, H.; Nagaoka, T. J. Electrochem.

Soc. 2016, 163, G166.

(30) Kinoshita, T.; Nguyen, D. Q.; Le, D. Q.; Ishiki, K.; Shiigi, H.; Nagaoka, T. Anal.

Chem. 2017, 89, 4680.

(31) Shiigi, H.; Kinoshita, T.; Fukuda, M.; Le, D. Q.; Nishino, T.; Nagaoka, T. Anal.

Chem. 2015, 87, 4042.

(32) Pettersson, L. A. A.; Johansson, T.; Carlsson, F. Synth. Met. 1999, 101, 198.

(33) Son, J. H.; Kim, G. H.; Kim, K. K.; Kim, S. H.; Kim, G. Y.; Lee, J. S. Bioresour.

Technol. 2003, 86, 215.

(34) Nguyen, D. Q.; Shan, X.; Saito, M.; Iwamoto, K.; Chen, Z.; Shiigi, H. Anal. Sci.

2019, 35, 763.

(35) Ashery, A.; Farag, A. A. M.; Shenashen, A. M. Synth. Met. 2012, 162, 1357.

(36) Rolfe, M. D.; Rice, C. J.; Lucchini, S.; Pin, C.; Thompson, A.; Alston, M.; Stringer,

M. F.; Betts, R. P.; Baranyi, J.; Peck, M. W.; Hinton, J. C. D. J. Bacteriol. 2012, 194, 686.

(37) Kolter, R.; Siegele, D. A.; Tormo, A. Annu. Rev. Microbiol. 1993, 47, 855.

(38) Nyström, T. Annu. Rev. Microbiol. 2004, 58, 161.

(39) Forrest, W. W.; Nyström, T. J. Bacteriol. 1965, 90, 1013.

(40) Lunt, S. Y.; Van der Heiden, M. G. Annu. Rev. Cell Dev. Biol. 2011, 27, 441.

(41) Arneborg, N.; Salskov-Iversen, A. S.; Mathiasen, T. E. Appl. Microbiol. Biotechnol.

1993, 39, 353.

(42) Yamamotoya, T.; Dose, H.; Tian, Z.; Fauré, A.; Toya, Y.; Honma, M.; Igarashi, K.;

Nakahigashi, K.; Soga, T.; Mori, H.; Matsuno, H. Biochim. Biophys. Acta, Proteins

Proteomics 2012, 1824, 1442.

31

Chapter III

Electrochemical Detection of Viable Bacterial Cells Using a Tetrazolium Salt

3.1. Introduction

Pathogenic bacteria cause infectious diseases and pose serious threats to human health.

To detect such bacteria, culture methods have been widely used as one of the most reliable

techniques.1−3 These techniques allow the detection of a single bacterium. However, the

main disadvantages are multistep assay that consists of pre-enrichment, selective

enrichment, isolation, and identification steps. Completion of all these phases requires at

least 16 h but can take as long as 48 h.4 Although recently developed all-in-one plating

systems have significantly reduced the labor-intensive operations, the assay still requires

up to 48 h for detection and enumeration of Escherichia coli and other fecal coliforms.5−7

To circumvent these problems, fast, simple, and sensitive techniques of bacterial

detection based on metabolism, antibody, and metal nanoparticles have been

developed.8−20

Metabolism-based techniques have advantages over other techniques in that neither

sophisticated instruments nor long processing times are required.21 Various types of

biosensors based on these techniques have been reported; Clark-type oxygen electrodes

were introduced to monitor microbial metabolism with a linear response ranging between

1.4 × 107 and 7.2 × 107 CFU mL–1.22 Different approaches have been reported using

intracellular enzymes;23−26 bacterial β-galactosidase hydrolyzes 4-aminophenyl-β-d-

galactopyranoside (APG) to electroactive 4-aminophenol, which allows the

amperometric detection of E. coli at a density as low as 4.5 × 102 CFU mL–1 after a 5.3 h

at 45 °C.27

As a different metabolism-based system, tetrazolium salts have become one of the

most widely used tools in cell biology.19,28−36 These salts, such as MTT and 5-cyano-2,3-

ditolyl tetrazolium chloride (CTC) (Scheme 3-1), are redox dyes that can rapidly penetrate

32

into intact biological cells, where they are enzymatically reduced to colored formazan

derivatives, which can be assessed by colorimetry or fluorometry.19 The salts do not

actually measure the number of viable cells, but rather respond to an integrated set of

enzymatic activities that are related to cell metabolism.19 Nevertheless, these salts have

been routinely used in biological and biomedical research,37 offering a robust measure of

viability that is demonstrated by the high correlation between viable cell numbers and

formazan signals.19 However, these tetrazolium salts have the disadvantage of requiring

prior solubilization of the generated formazan for use in a colorimetric assay.33,38

Alternatively, water-soluble salts, such as 2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-5-

[(phenyl-amino) carbonyl]-2H-tetrazolium hydroxide (XTT) and 2-(2-methoxy-4-

nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfo-phenyl)-2H-tetrazolium, monosodium salt

(WST-8), have recently been used to circumvent this problem.29,31,32,39

Scheme 3-1. Chemical structures of tetrazolium salts, MTT and CTC, and their reduction

forms (formazan).

However, in electrochemical assays, I have found that the insolubility of formazan

(FORMH) can be beneficially exploited as a surface-confined redox reaction after the in-

situ desiccation (thermal lysis) of microbial cells deposited on an indium-tin-oxide (ITO)

33

electrode. Desiccation transfers formazan from the microbial cells to the electrode sur-

face, and this formazan is consequently maximally concentrated on the electrode by

adsorption. In this paper, I apply this novel yet simple technique to a sensitive microbial

assay. Thermal lysis was performed by disrupting microbial membranes at high

temperatures, and was successfully applied for the extraction of DNA.40 Microfluidic

devices integrating the thermal lysis of microbes have been engineered for use in

downstream DNA analysis.41 Apart from DNA analysis, the in-situ thermal lysis of

bacterial cells make it possible to deposite bacterial cells on an ITO electrode, can be

effectively used for the detection of cellular hydrophobic molecules, such as ubiquinone

and menaquinone, in combination with electroanalysis.42,43 As this technique only detects

hydrophobic molecules adsorbed on the electrode, highly selective detection is made

possible by avoiding interference from hydrophilic redox molecules that are present in

microbial cells. Using this technique, I have reported that cellular ubiquinone and

menaquinone are quantified in a straightforward manner and have applied this technique

to profiling the modes of respiration for some facultative anaerobes such as E. coli and

Shewanella oneidensis.42,43 In this section, I demonstrate that this in-situ lysis adsorption

technique can also be applied to the detection of metabolically generated hydrophobic

indicator molecules, other than isoprenoid quinones.

3.2. Experimental

3.2.1. Microbe culture and chemicals

Inocula used for the cultures were obtained from single colonies on agar plates. All

culture media were autoclaved before use. The microbial species used in this section were

purchased from the Biological Resource Center (nite), Japan. In this paper, E. coli K-12

strain is denoted as E. coli. Cultures were incubated aerobically at 30°C for >18 h. All the

microbes used in this section were cultured in nutrient broth (abbreviated as NB; E-MC35,

EIKEN CHEMICAL Co. Ltd., Japan). The OD600 value was correlated to the bacterial

colony count with a Petrifilm count-plate AC (3M Health Care); i.e. 1.00 OD600 ≡ 1.05 ×

109, 1.90 × 108, 1.36 × 109, 3.80 × 108 CFU mL－1 for E. coli, Pseudomonas aeruginosa

34

(P. aeruginosa), Salmonella enterica (S. enterica), and Staphylococcus aureus (S. aureus),

respectively. MTT was purchased from DOJINDO, Japan. All other chemicals were of

analytical grade and were used as received. Milli-Q quality water was autoclaved and

used throughout.

3.2.2. Procedures and apparatus.

All microbial experiments were performed under strictly sterile conditions. An

Ag|AgCl| saturated KCl| electrode, and a platinum coil electrode (4 mm × 1.3 cm), were

used as the reference and counter electrodes, respectively, throughout this study. An ITO-

coated glass strip with dimensions of 26 × 77 mm2, obtained from Kinoene Optics (Japan),

was used as the working electrode (10 square). Its surface was covered with a UV-

curing resin film using a Roland DG LEF12 inkjet printer, with the exception of a circular

area (2.0 mm diameter) used as the working electrode. The cyclic voltammograms (CVs)

were recorded with an ALS CHi842B Electrochemical Analyzer at ambient temperature

(25±1°C). Scanning electron microscope (SEM) images were obtained with an TM-3030

instrument (Hitachi, Japan). Dark-field observations were performed using an optical

microscope (ECLIPSE Ni, Nikon, Japan) with a dark-field condenser, a 100 W halogen

lamp, and a camera equipped with a charge-coupled device (DS-Ri1, Nikon, Japan).

Samples for analysis were prepared in the following way (Scheme 3-2): The bacterial

culture was diluted with NB to prepare a suspension with a microbial density ranging

from 100 to 108 CFU mL－1. This suspension of 950 µL was transferred to a 1.5 mL

microtube and incubated for 1 h at 37 °C after the addition of 50 µL of 10 mM MTT (1

M ≡ 1 mol L–1). During the incubation, no bacterial growth occurred. The incubated

suspension was then centrifuged for 10 min at 2,000 ×g, and the supernatant was removed.

Sterile water of 1,000 µL was then added to the residue and the microtube was centrifuged

again. Subsequently, the supernatant was removed and 10 µL of water was added to rinse

the inner wall of the microtube. The suspension, including the residue, was well mixed.

Then, a 5.0 µL aliquot was dispensed on the ITO electrode, which was then completely

dried with a heat gun for 1 min. The CV of the dried electrode was measured at 60 mV s–

1 in nitrogen-saturated phosphate buffer (pH: 7.0).

35

This technique was applied for the determination of microbial density in a fluid

fertilizer, which had been continuously circulated in the flow channel of the vegetable

cultivation plan, by the standard addition technique. This fluid (100 µL) was combined

with 900 µL of NB and 50 µL of 10 mM MTT, which was followed by the procedures as

stated previously. To prepare a calibration plot, E. coli ranging from 102 to 107 CFU mL–

1, was dispersed in the fluid sample before incubation. The density of the common

bacteria in this fluid was confirmed, using a Petrifilm count-plate, as 3.0×105 CFU mL－

1. The densities of coliforms and E. coli were estimated as 21 and 0 CFU mL－1,

respectively, with Sanita-kun sheet-medium for E. coli & Coliforms, JNC (Japan),

respectively.

Scheme 3-2. An outline of the experimental procedures: (A) Uptake of MTT, (B)

microbial reduction of MTT to formazan, (C) application of the microbe on an ITO

electrode, and (D) in-situ lysis-adsorption followed by voltammetric measurement.

3.3. Results and Discussion

Figure 3-1 shows the CVs of the CTC and MTT dyes. A solution containing one of

the two dissolved dyes was applied to an ITO electrode, dried, and subjected to

voltammetric analysis.

36

Figure 3-1. CVs of (A) 1.6×10–6 mol cm–2 CTC and (B) 1.6×10–8 mol cm–2 MTT applied

to the ITO electrodes.

Electrochemical responses were attributed to the redox reaction of these dyes

absorbed on the electrode, since there are no redox-active substances in the test solution.

Oxidized-form CTC and MTT dyes were transformed into reduced-form FORMH

precipitates from a positive to negative potential sweep, while FORMH was re-oxidized

from negative to positive potential scan. Removal of dissolved oxygen by nitrogen-

bubbling allows us to obtain current responses of MTT and CTC dyes, since dissolved

oxygen induces a narrow potential window on account of oxygen reduction occurring

below –0.3 V (vs. Ag|AgCl).44 These dyes have the ability to change color depending on

their redox states and have been frequently used for the colorimetric measurement of

microbial metabolic activity. The basic difference between the molecular structures of

CTC and MTT occurs at the 5-position of the tetrazolium ring, and the higher

hydrophobicity of MTT is presumably responsible for the >100-fold increase in

magnitude of the redox peaks that led us to choose MTT for the following study.

Previous studies suggested that the microbial isoprenoid-quinones (ubiquinone and

menaquinone) present in the cytoplasmic membrane can be transferred to the electrode

37

surface by in-situ desiccation of microbes deposited on the electrode.42,43 Desiccation

destructs the outer and cytoplasmic membranes and causes these hydrophobic quinones

to be adsorbed on the ITO electrode. Therefore, the detection of quinones can be

effectively coupled with thermal lysis. Based on this protocol, I have attempted the

quantification of the microbially formed FORMH, which is transferred from the microbe

to the electrode upon desiccation and adsorbed to the electrode surface, as in the case of

isoprenoid quinones.

Figure 3-2. CVs of FORMH generated by incubation for 1 h at 37°C in NB with 0.50

mM MTT and (a) E. coli at densities of 5.5×106, (b) 5.5×105, (c) 5.5×103, (d) 5.5×102, or

(e) 0 CFU mL–1. The inset shows the amplified voltammograms for the Ia region.

Figure 3-2 shows the CVs of FORMH generated by E. coli. Insoluble FORMH,

enzymatically formed in the microbial cells, was transferred to an ITO electrode by

thermal-lysis, and CVs were recorded in a nitrogen-saturated phosphate buffer (pH: 7.0)

from the negative to positive potential scan. The CVs scanned from –0.7 V exhibited high

oxidation peaks at +0.1 V and +0.75 V. The accumulation of insoluble FORMH

precipitate on the ITO electrode allows us to obtain larger current responses than that of

soluble MTT. The oxidation peak at +0.1 V was used for the determination of the

microbial density in this work. To confirm that this symmetric peak arose from an

38

adsorption event, CV experiments were conducted by varying the scan rate. The peak

current was found to be linearly proportional to the scan rate, indicating that the oxidation

was ascribed to a surface-confined process. As the electrode was preequilibrated in the

phosphate buffer, hydrophilic substances, which had been discharged from the microbes

during thermal lysis, were effectively removed from the electrode surface before potential

scanning. In addition, no peaks were found for redox protein molecules.42,43 Thus, the

obtained responses were highly selective to FORMH, as seen in Figure 3-2. In addition,

the small background current of the ITO electrode allowed to acquire voltammograms

with a better S/N ratio. Indeed, its base current was much lower than that for other

electrode materials, such as glassy carbon and gold, and the low surface roughness of the

ITO-coated glass electrode would be responsible for the small current.42,43

To acquire peak assignments, CVs were recorded by varying the initial and switching

potentials, which resulted in the same assignments as those reported by Marques et al for

MTT adsorbed on a pyrolytic graphite electrode.45 The peak separation of IIc/IIa in this

study was, however, larger than those observed, and might be dependent on the electrode

material and immobilization conditions. The oxidation of FORMH to MTT includes the

stepwise removal of two electrons and one proton and the first oxidation peak, Ia, likely

represents the formation of an MTT radical. However, Umemoto et al proposed a different

redox mechanism for 2,3,5-triphenyltetrazolium chloride in dimethylsulfoxide (DMSO)

based on the disproportionation of its radical and the paired Ia and IIc peaks.46 Moreover,

water-soluble tetrazolium salts exhibited different electrochemical responses as

demonstrated by our results.47,48 Therefore, it is difficult to discuss the precise redox

mechanism due to the limited information available for the electrochemistry of the

tetrazolium salts.45-48 Aside from uncertainty in the redox mechanism, it was observed

that peak IIc was much smaller than the other three peaks. A diffusional process, rather

than adsorption, might be responsible for the small trailing peak.

It is noteworthy that peak Ia can detect a microbial density less than 5.5×102 CFU mL–

1, as seen in the inset. The control signal (curve e) in the inset arose from the MTT reagent

that was not removed at the centrifugation stage. MTT was reduced to FORMH, while

39

the electrode was poised at –0.7 V before starting the scan. As mentioned earlier, the

lysed microbes show CV peaks arising from the ubiquinone and menaquinone redox.42,43

However, these peaks were only detectable for high-density microbial suspensions ( 107

CFU mL–1) under the present conditions, and the quinone peaks did not interfere with the

observation of FORMH.

The nutrient medium, NB, helped accelerate the microbial reduction of MTT. E. coli

incubated in NB showed a 104-fold electrochemical response compared to that for saline

solution. Stowe et al reported that FORMH produced by E. coli was inhibited at high

MTT concentration ( 0.648 mM) due to the toxicity of the tetrazolium salt itself.38 In

contrast, lower MTT concentrations require longer incubation time. Considering these

two factors, i.e., sensitivity and rapidity, I adopted that MTT concentration of 0.50 mM

in the following experiments.

Figure 3-3. (A) CVs of FORMH generated by incubation for 0, 0.5, 1.0, 2.0, and 3.5 h in

NB medium with 0.50 mM MTT and E. coli (2.0×106 CFU mL–1). CVs were recorded at

a scan rate of 60 mV s–1 in a nitrogen-saturated phosphate buffer (pH 7.0). (B)

Dependence of oxidation peak current in Ia reagion on incubation time.

The peak current of Ia increased with incubation time during early stage of incubation.

With an E. coli density of 2.0 × 106 CFU mL–1 and an MTT concentration of 0.50 mM,

the current reached 40%, 96%, and 100% at 0.5, 1.0, and 3.5 h (Figure 3-3), respectively.

At lower densities (7.0 × 105 and 6.0 × 103 CFU mL–1), the current was virtually the same

for incubation times between 1 and 5 h. From these results, I have concluded that an

incubation for 1 h at 37°C is sufficient for microbes to reduce the MTT uptake.

40

A metabolism-based assay usually requires a longer incubation time with a decrease

in microbial density. APG hydrolysis by E. coli required incubation times of 2.3 and 7.1

h for the detection of 1.0 × 106 and 1 CFU mL–1. In order to increase absorbance by 0.5,

WST-8 required incubation times of 1 and 4 h for E. coli with a density of 1.3 × 107 and

1.3 × 105 CFU mL–1, respectively.32 These incubation times are related to the

accumulation of the soluble indicator molecules that were metabolically formed. In

contrast, it was reported that MTT reduction in an LB-glycerol liquid medium was

completed in 20 min, independent of E. coli density,37 and my data also showed little

dependence of the CV response on the incubation time exceeding 1 h. As the growth of

the FORMH crystal compromised cell metabolism, the CV signal intensity leveled out

after 1 h. For this reason, microbial suspensions in this study were incubated for 1 h.

Figure 3-4. Dependence of oxidation peak current Ia on microbial density. Microbial

species: (◆) S. enterica, (■) P. aeruginosa, (●) E. coli, and (▲) S. aureus. Scan rate, 60

mV s–1. Inset: amplified diagram. Error bars indicate standard deviation (n=3).

As seen in Figure 3-4, four microbial species showed similar dependence on microbial

density. The onset of the rise in current was almost the same among these microbes seen

in the inset. As mentioned, small current responses also occurred in bacterial free solution

due to residue of MTT reagents. The current responses in bacteria-free solution was

41

obtained as 0.50 ± 0.13 μA cm–1, which was almost identical with the signal obtained

from the smallest amounts of bacterial suspension (< 10 CFU mL–1). However, the

response (1.1 ± 0.23 μA cm–1 of average among four microbial species) obtained in

bacterial suspension at around 5.3 × 101 CFU mL–1 was obviously higher than those in

bacteria-free solution. In addition, there is not very good correlation between the current

density and the number of bacteria (correlation coefficient: R2=0.7803). The current

responses depend strongly on the bacterial viability and metabolic activity rather than the

number of cells in the suspension. It is not possible to control the activity of individual

cells in the suspension while the viability can be estimated around 95 % through the

experiments. Therefore, I supposed that the results include variations in metabolic activity.

The limit of detection (LOD) was estimated by the three-sigma method as 2.8 × 101 CFU

mL–1, the value of which was evaluated from the intersection of the calibration curve with

the sum of the average of the blank values (Ave.) and three times the standard

division.50,51 It is noteworthy that the present technique achieved up to a 10,000-fold

increase in sensitivity over MTT colorimetric assays, of which the LOD was estimated as

105 CFU mL–1 higher, and the 1 h incubation time was much shortened from that

employed by metabolism-based systems that often required incubation times >4 h.31-

33,38,49 The sensitivity enhancement is primarily due to the insolubility of FORMH as

follows: Its insolubility makes it possible (i) to concentrate FORMH on the electrode to

the maximal degree through desiccation and adsorption, (ii) to immobilize FORMH on

the electrode to prevent it from diffusing away in solution during the CV measurement,

and (iii) to observe an intense adsorption CV peak. The current responses increased

sharply at around 105 CFU mL–1. However, marked diminutions in the current peaks were

found for all microbes above 106 CFU mL–1 that were attributed to the deposition of

sizeable FORMH crystals on the electrode.

Figure 3-5A shows the SEM images of the electrodes bearing E. coli. Bacterial cells

were found as rod-like crystals on the electrode at 105 CFU mL–1 (Figure 3-5Ab), while

the crystals were observed as small pieces in the lower bacterial concentration region at

less than 104 CFU mL–1 due to the dissociation of bacterial cells by thermal lysis (Figure

3-5Aa). This was supported by the dark-field microscopic observation revealing that

42

sphere-like FORMH crystals were generated around the bacterial surface before thermal

lysis (Figure 3-5Bb). Long crystals of FORMH could be observed at higher microbial

densities (>106 CFU mL–1) and many crystals grew much bigger than the microbial bodies

after thermal lysis, and the inset in Figure 3-5Ac reveals the crystals were aciculate when

they were small.

Figure 3-5. (A) SEM images of the electrode surfaces after thermal lysis with different

E. coli densities. E. coli densities are presented in CFU mL–1 at (a) 5.5 × 104, (b) 5.7 ×

105, (c) 5.5 × 106, and (d) 5.5 × 107. The images were acquired after voltammetric

experimentation. Scare bars represent 30 μm. (B) Dark-field images of E. coli cells (8.0

× 109 CFU mL–1) without thermal lysis (a) before and (b) after incubation with 0.50 mM

MTT for 1 h at 30 °C. Acquisition time of dark-field images is 400 ms. Scare bars

represent 5 μm.

This extreme morphological change depended strongly on the amount and density of

FORMH produced by the cells. Consequently, the decrease in the current density above

106 CFU mL–1 can be explained by the low electroconductivity of the crystal. Below this

43

density, a thin electroactive FORMH layer was presumably formed on the electrode.

However, the crystal growth made electron abstraction from its inside difficult and

eventually reduced the current intensity. When microbes >106 CFU mL–1 were dispensed

on the electrode, the violet color of FORMH was easily recognized by the naked eye.

Thus, the overloading, which leads to a false quantification, is easily detected by

observing the color of the electrode surface before voltammetric measurement. On the

contrary, I supposed that the metabolic ability depends on the microbial density since it

is well-known that high concentration of bacterial cells induces to unique metabolic

activity with quorum sensing.52

To understand how the crystals grew upon incubation, the absorption spectra of E.

coli were observed. When the incubated microbe was dispersed in water (8.0 × 109 CFU

mL–1), the spectrum showed a band maximum at 567 nm. The band intensity of the

suspension, after passage through a 0.2 μm filter, diminished to 11% compared to that of

the unfiltered suspension. In Figure 3-5Bb, sphere-like FORMH crystals were generated

around bacterial surface after incubation with MTT, as previously reported.37,53 These

results suggest that the FORMH crystals mainly present in/on the microbial cells.

Next, DMSO was added to the microbes to dissolve the crystals, and the spectra were

observed. The solution had a purple color with an absorption maximum of 550 nm, and

the absorbance remained the same after passage through the filter. This suggests that

FORMH was completely extracted from the cells and solubilized in DMSO. The results

in the aqueous and nonaqueous solvents indicated that ∼90% of the FORMH crystals

present in/on the cells and were transferred to the electrode surface by the destruction of

the cell wall during desiccation. The rest of the crystals (∼10%) were smaller than 0.2 μm

and probably aciculate.52 These crystals easily penetrated the cell wall, resulting in their

ejection into the solution.

As seen in Figure 3-5Ad, many crystals grew to a length of ∼10 μm on the electrode,

and the growth presumably occurred during desiccation. At the beginning of the drying

procedure, the crystals deposited on the electrode surface could be temporarily dissolved

by an increase in temperature. The successive loss of water, due to evaporation, then

44

caused dissolved FORMH to recrystallize, making the crystals grow up to a length of ∼

10 μm. Before thermal lysis dissociation, FORMH crystals were obtained as a sphere-like

structure mostly bound to the bacterial cell surface (Figure 3-5Bb). Accordingly, the large

needle crystals, which were separated from the microbes (Figure 3-5Ad), also suggest

that they grew on the electrode during the desiccation. Figure 3-5Ad also reveals that the

microbes discharged their internal fluid by thermal lysis, as seen as the gray area.

Figure 3-6. E. coli (1.0 × 108 CFU mL–1) incubated with 0.24 mM MTT for 2.5 h. Dead

microbes were prepared by autoclaving for 15 min at 120 °C before incubation with MTT.

Figure 3-6 shows the CV responses for viable and nonviable E. coli cells incubated

with MTT. MTT is a redox dye, which is reduced to FORMH in viable cells with

NAD(P)H-dependent oxidoreductases and dehydrogenases. Indeed, no response was

observed for the heat-killed cells, as seen in Figure 3-6. Previous report suggested that

high microbial density (≥109 cells cm–2) on the electrode increases the background current

due to bacterial cell insulation.42 Therefore, it was difficult to observe a faradaic current

of residual MTT in nonviable E. coli in the voltammogram. Cell proliferation and viability

assays are of particular importance for routine applications in bacterial biology, so the

detection of viable microbes is of major concern for many applications such as testing

antibiotic effects against microbes.19,31,35,38 If this technique were to be used as an

alternative to colony counting, it could also be employed for viability and activity

detection.

45

Figure 3-7. Determination of microbial density in fertilizer fluid with the standard

addition of E. coli. Error bars indicate standard deviation (n = 6).

Figure 3-7 shows the application of this technique to fertilizer fluid circulated through

the flow channels for cultivating vegetables. Microorganisms in fertilizer fluid play an

important role to provide indispensable nutrients such as nitrogen and phosphorus for

plants. Therefore, microbial density in fertilizer fluid is one of the critical factors to

cultivate plants. E. coli was added to the fluid to prepare this calibration plot, although

the fluid did not contain E. coli (0 CFU mL–1). The density of viable microbes, measured

with a Petrifilm, was 4.8 ± 0.2 × 105 CFU mL–1. The fluid was diluted by a factor of 10.5

to incubate the microbes with the MTT/NB solution. The obtained value by the

electrochemical method was estimated as 4.5 ± 0.2 × 104 CFU mL–1, which corresponds

to 4.6 ± 0.4 × 105 CFU mL–1, demonstrating an excellent agreement with that of the count-

plate technique. There is no effect of sample matrix caused by adsorbent and reductant

components. The MTT assay usually has low specificity for microbial species, and the

responses in the low-density region were not significantly different among the microbes

(Figure 3-4). Consequently, the addition of an indifferent microbe to the fluid did not

result in a significant error. The total assay time, including all the steps such as sampling,

incubation, and electrochemical measurement required by this technique, is

approximately 1.5 h, which is much shorter than that required for cultivation and other

metabolism-based techniques (e.g., > 24 h for Petrifilm).

46

3.4. Conclusion

In this work, the electrochemical detection of formazan was effectively coupled with

the in situ thermal lysis of microbes. The presented technique was capable of detecting

viable microbes at densities of above 2.8 × 101 CFU mL–1. The results of this study

indicate that the present technique has sensitivity up to 10,000-fold higher compared to

that of the formazan colorimetric method and requires an incubation time of only 1 h,

which is approximately 1/4 of that required for other metabolism-based techniques. The

higher sensitivity is mainly ascribed to the reduction of microbially produced formazan

to an extremely small volume (film) through evaporation, coupled with the baseline-

separated intense adsorption peak. My technique is useful for trace samples and is

therefore applicable to determine the number of microorganisms not only for liquids such

as tears, saliva, and sweat but also for samples obtained by wiping off from solids such

as skin, food, and public objects to prevent infectious diseases and food poisoning and

control quality of foods.

47

REFERENCES

(1) Yoshida, S. Toda's New biacteriology, 34th ed.; Nanzando Co. Ltd.: Tokyo, 2013.

(2) Madigan, M. T.; Martinko, J. M.; Parker, J. Brock Biology of Microorganisms, 9th

ed.; Prentice Hall, INC., 2000.

(3) Sobsey, M. D. In Identifying Future Drinking Water Contaminants, Macdonald, J. A.;

Gibson, M.; Swartz, K. A., Eds.; National Academy Press: Washington, D.C., 1999.

(4) Ivnitski, D.; Abdel-Hamid, I.; Atanasov, P.; Wilkins, E.; Stricker, S. Electroanalysis.

2000, 12, 317.

(5) Russell, S. M. J. Food Prot. 2000, 63, 1179.

(6) Schraft, H.; Watterworth, L. A. J. Microbiol. Methods 2005, 60, 335.

(7) Vail, J. H.; Morgan, R.; Merino, C. R.; Gonzales, F.; Miller, R.; Ram, J. L. J. Environ.

Qual. 2003, 32, 368.

(8) Zourob, M.; Elwary, S.; Turner, A. P. F. In Principles of Bacterial Detection:

Biosensors, Recognition Receptors and Microsystems; Springer-Verlag: New York, 2008.

(9) Brewster, J. D.; Gehring, A. G.; Mazenko, R. S.; Van Houten, L. J.; Crawford, C. J.

Anal. Chem. 1996, 68, 4153.

(10) Abdel-Hamid, I.; Ivnitski, D.; Atanasov, P.; Wilkins, E. Biosens. Bioelectron. 1999,

14, 309.

(11) Ivnitski, D.; Abdel-Hamid, I.; Atanasov, P.; Wilkins, E. Biosens. Bioelectron. 1999,

14, 599.

(12) Shiigi, H.; Kinoshita, T.; Fukuda, M.; Le, D. Q.; Nishino, T.; Nagaoka, T. Anal.

Chem. 2015, 87, 4042.

(13) Kinoshita, T.; Kiso, K.; Le, D. Q.; Shiigi, H.; Nagaoka, T. Anal. Sci. 2016, 32, 301.

(14) Kinoshita, T.; Nguyen, D. Q.; Le, D. Q.; Ishiki, K.; Shiigi, H.; Nagaoka, T. Anal.

48

Chem. 2017, 89, 4680.

(15) Shan, X.; Yamauchi, T.; Yamamoto, Y.; Niyomdecha, S.; Ishiki, K.; Le, D. Q.; Shiigi,

H.; Nagaoka, T. Chem. Commun. 2017, 53, 3890.

(16) Shan, X.; Yamauchi, T.; Yamamoto, Y.; Shiigi, H.; Nagaoka, T. Analyst 2018, 143,

1568.

(17) Mandal, P. K.; Biswas, A. K.; Choi, K.; Pal, U. K. Am. J. Food Technol. 2011, 6,

87.

(18) Botes, M.; De Kwaadsteniet, M.; Cloete, T. E. Anal. Bioanal. Chem. 2013, 405, 91.

(19) Berridge, M. V.; Herst, P. M.; Tan, A. S. Biotechnol. Ann. Rev., 2005, 11, 127.

(20) Sepunaru, L.; Tschulik, K.; McAuley, C. B.; Gavish, R.; Compton, R. G. Biomater.

Sci. 2015, 3, 816.

(21) Palchetti, I.; Mascini, M. Anal. Bioanal. Chem. 2008, 391, 455.

(22) Suzuki, H.; Tamiya, E.; Karube, I. Electroanalysis 1991, 3, 53.

(23) Palcheti, I.; Mascini, M. In Principles of Bacterial Detection: Biosensors,

Recognition Receptors and Microsystems, Zourob, M.; Elwary, S.; Turner, A. P. F., Eds.;

Springer-Verlag: New York, 2008.

(24) Serra, B.; Morales, M. D.; Zhang, J.; Reviejo, A. J.; Hall, E. H.; Pingarron, J. M.

Anal. Chem. 2005, 77, 8115.

(25) Mulchandani, P.; Hangarter, C. M.; Lei, Y.; Chen, W.; Mul-chandani, A. Biosens.

Bioelectron. 2005, 21, 523.

(26) Togo, C. A.; Wutor, V. C.; Limson, J. L.; Pletschke, B. I. Biotechnol. Lett. 2007, 29,

531.

(27) Pérez, F.; Tryland, I.; Mascini, M.; Fiksdal, L. Anal. Chim. Acta 2001, 427, 149.

(28) Mosmann, T. J. Immunol. Methods 1983, 65, 55.

49

(29) Meletiadis, J.; Mouton, J. W.; Meis, J. F. G. M.; Bouman, B. A.; Donnelly, J. P.;

Verweij, P. E.; Network, E. J. Clin. Microbiol. 2001, 39, 3402.

(30) Bhupathiraju, V. K.; Hernandez, M.; Landfear, D.; Alvarez-Cohen, L. J. Microbiol.

Methods 1999, 37, 231.

(31) Tunney, M. M.; Ramage, G.; Field, T. R.; Moriarty, T. F.; Storey, D. G. Antimicrob.

Agents Chemother. 2004, 48, 1879.

(32) Tsukatani, T.; Suenaga, H.; Higuchi, T.; Akao, T.; Ishiyama, M.; Ezoe, K.;

Matsumoto, K. J. Microbiol. Methods 2008, 75, 109.

(33) Denizot, F.; Lang, R. J. Immunol. Methods 1986, 89, 271.

(34) Carmichael, J.; Degraff, W. G.; Gazdar, A. F.; Minna, J. D.; Mitchell, J. B. Cancer

Res. 1987, 47, 936.

(35) Ukeda, H.; Goto, Y.; Sawamura, M.; Kusunose, H.; Kamei, T. Nippon Shokuhin

Kagaku Kogaku Kaishi 1995, 42, 627.

(36) Ukeda, H.; Maeda, S.; Ishii, T.; Sawamura, M. Anal. Biochem. 1997, 251, 206.

(37) Wang, H.; Cheng, H.; Wang, F.; Wei, D.; Wang, X. J. Microbiol. Methods 2010, 82,

330.

(38) Stowe, R. P.; Koenig, D. W.; Mishra, S. K.; Pierson, D. L. J. Microbiol. Methods

1995, 22, 283–292.

(39) Tsukatani, T. Nippon Shokuhin Kagaku Kogaku Kaishi 2015, 62, 321.

(40) Bao, N.; Lu, C. In Principles of Bacterial Detection: Biosen-sors, Recognition

Receptors and Microsystems; Zourob, M.; Elwary, S.; Turner, A. P. F., Eds.; Springer-

Verlag: New York, 2008.

(41) Privorotskaya, N.; Liu, Y. S.; Lee, J.; Zeng, H.; Carlisle, J. A.; Radadia, A.; Millet,

L.; Bashir, R.; King, W. P. Lab Chip 2010, 10, 1135.

(42) Le, D. Q.; Morishita, A.; Tokonami, S.; Nishino, T.; Shiigi, H.; Miyake, M.; Nagaoka,

50

T. Anal. Chem. 2015, 87, 8416.

(43) Morishita, A.; Higashimae, S.; Nomoto, A.; Shiigi, H.; Nagaoka, T. J. Electrochem.

Soc. 2016, 163, G166.

(44) Wang, Z.; Deng, H.; Chen, L.; Xiao, Y.; Zhao, F. Bioresour. Technol. 2013, 132,

387.

(45) Marques, E. P.; Zhang, J. J.; Tse, Y. H.; Metcalfe, R. A.; Pietro, W. J.; Lever, A. B.

P. J. Electroanal. Chem. 1995, 395, 133.

(46) Umemoto, K. Bull. Chem. Soc. Jpn. 1989, 62, 3783.

(47) Oritani, T.; Fukuhara, N.; Okajima, T.; Kitamura, F.; Ohsaka, T. Inorg. Chim. Acta

2004, 357, 436.

(48) Nissim, R.; Compton, R. G. ChemElectroChem 2014, 1, 763.

(49) Stiefel, P.; Schneider, J.; Amberg, C; Maniura-Weber K.; Ren, Q. Sci. Rep. 2016, 6,

39635.

(50) Olanrewaju, A. O.; Ng A.; Martin, P. D.; Robillard A.; Juncker D. Anal. Chem. 2017,

89, 6846.

(51) Raut N.; Pasini P.; Daunert S. Anal. Chem. 2013, 85, 9604.

(52) Davies, G. D.; Parsek, M. R.; Pearson, J. P.; Iglewski, B. H.; Consterton, J. W.;

Greenberg, E. P. Science, 1998, 280, 295.

(53) van den Berg, B. M. Open Vet. J. 2015, 5, 58.

51

Chapter IV

Investigation of Precious Metal-ion Reduction by Shewanella oneidensis MR-1

4.1. Introduction

Shewanella oneidensis MR-1 is a facultative anaerobic bacterium that is known to

extracellularly transfer electrons generated during metabolism to various soluble or

insoluble substances as terminal electron acceptors under anaerobic conditions. Many

applications using extracellular electron transfer have been reported concerning

environmental and energy-creation biotechnologies, such as for the collection of precious

metal ions,1-3 radionuclides,4 and biocatalyst in microbial fuel cells.5,6 The electron

carriers are mainly c-type cytochrome proteins, which are arranged on the bacterial

inner/outer membrane. Electrons in the menaquinol pool flow to CymA at the inner

membrane, and then to MtrA in the periplasmic space. From MtrA, electrons are passed

through MtrB, and finally to MtrC and OmcA in the outer membrane.7-14 The electrons

transported by this pathway are finally exposed to the extracellular environment and

contact terminal electron acceptors. Primarily, a final redox reaction that proceeds through

MtrC and OmcA have important roles in the reduction of various metal ions, including

Pd(IV), U(VI), Ag(I), and Fe(III).1,13-15 Previous studies showed that the reduction of

Au(III) ion to gold nanoparticles (Au NPs) was successfully performed using S.

oneidensis, and that Au NPs were produced on the bacterial surface.2,16 However, the

contribution of the bacterial surface structure to the production of Au NPs remained

unclear.

Metal NPs are well-known to enhance light absorption and scattering in a specific

wavelength region, based on their localized surface plasmon resonance (LSPR).

Therefore, it is possible to observe metal NPs below the theoretical resolution of an

optical microscope (~200 nm) owing to their enhanced light-scattering intensity. The

light-scattering spectrum of the NPs depends strongly on not only the metal species, but

also on the size and aggregation/dispersion state.

52

In this study, I tracked the formation process of Au NPs by using electron/dark-field

microscopy, zeta potential analysis, and spectrometry and evaluated the roles of

membrane proteins and extracellular polysaccharide (EPS) in the reduction of the Au (III)

ion. I have also proposed a method for optical elemental analysis of metal species using

S. oneidensis, focusing on the light scattering properties of metal NPs on bacterial surface.

4.2. Experimental

4.2.1. Bacterial culture and purification

The bacterial strain S. oneidensis MR-1 was purchased from American Type Culture

Collection (Manassas, VA, USA). A strain of S. oneidensis was cultured in an agar growth

medium (E-MC35, Eiken Chemical Co., Tokyo, Japan) at 303 K for 24 h. Colonies were

suspended in a liquid medium (Nutrient Broth, Eiken Chemical Co.) and incubated at 303

K for 24 h with shaking. After cultivation, the cells were obtained as a precipitate

following centrifugation at 8200 × g for 10 min, followed by resuspension in fresh

phosphate-buffered saline (pH 7.4). This procedure was repeated three times to obtain

purified cells. The resulting suspension (2.0 × 108 cells mL–1) was used for following

experiments.

4.2.2. Metal-ion reduction

The suspension containing sodium formate (0.10 M) and a metal source, such as

palladium(II) chloride, hexachloroauric(III) acid, hexachloroplatinic(IV) acid, copper(II)

sulfate, copper(II) chloride, or nickel(II) chloride were added to the bacterial suspension

(20 mL), was incubated under a nitrogen-saturated atmosphere at 298 K.

4.2.3. Apparatus

Scanning electron microscope (SEM) images were obtained with an TM-3030

instrument (Hitachi, Japan). Transmission Electron Microscope (TEM) image of the

bacterial mixture was obtained with a JEM 2000FXII (Jeol, Japan) at an accelerating

voltage of 200 kV. Zeta potential was measured with a zeta-potential and particle size

analyzer (ELSZ-2Plus, Otsuka Electronics, Japan). UV-vis absorption spectra of the

53

bacterial mixture were measured with a UV-vis spectrometer (UV-3100PC, Shimadzu,

Japan).

4.2.4. Dark-field observation and measurement of light-scattering spectra

Samples were prepared in following way: 5 μL of bacterial mixture was pipetted onto

a glass slide and dried at room temperature for 20 min. Dark-field observation was

performed using an optical microscope (ECLIPSE Ni, Nikon, Japan) with a dark-field

condenser, a 100 W halogen lamp, and a camera equipped with a charge-coupled device

(DS-Ri1, Nikon, Japan). Light-scattering spectra were measured using a miniature grating

spectrometer (USB4000, Ocean Optics, FL), which was connected to the microscope

using an optical fiber (core diameter, 400 μm).

4.3. Results and Discussion

The color of the suspension changed from light yellow (tetrachloroauric acid color)

to light red purple over a period of 3 h, and gradually became dark reddish-purple because

of a localized surface plasmon resonance (LSPR) of the Au NPs.17,18 In accordance with

the color change, the UV-vis spectra revealed no peaks at the early stage of incubation;

however, a peak at approximately 550 nm gradually became intense, as shown in Figure.

4-1A.

Because bacteria have no absorption peaks other than that at 260 nm, derived from

their nucleic acids, the absorption at 550 nm was assigned to the LSPR of the Au NPs

generated. In addition, an increase in the intensity of the LSPR induced a red-shift from

550 to 600 nm, which depended on the size and/or dispersion state of the Au NPs.19,20

This indicated that the bacterial cells consumed formate to generate electrons, which

reduced aurate to Au NPs. It was confirmed that formate did not function as a reducing

agent, since there was no change in the color of the solution without the bacterial cells.21

Moreover, absorption observed over a wide wavelength range as the baseline became

larger with the incubation time because of the increase in cell numbers, leading to a more

54

turbid bacterial suspension. The increase in the bacterial cell numbers with the incubation

time was verified by colony counting (6.0 × 108 cells mL–1 at 10 h).

Figure 4-1. (A) UV-vis spectra of bacterial suspensions containing 1.0 mM

hexachloroauric (III) acid during incubation and (B) SEM images of cells before and after

incubation for 24 h.

SEM images indicated a large difference in contrast between before and after

incubation, as shown Figure 4-1B. The bacterial cells appeared as dark-grey rods before

incubation because of their insulation, while after incubation many bright rods were

observed on the substrate. According to the energy-dispersive x-ray spectrometry results,

the bacterial cells were confirmed to be coated with Au elements, and therefore their

conductivities increased by forming Au NPs.

To investigate changes to the cell surface along with the formation of Au NPs, a zeta

potential measurement was carried out. Gram-negative bacteria, such as Escherichia coli,

Pseudomonas aeruginosa, and Salmonella enterica, have a negatively charged surface

derived from phosphate and carboxylate groups including lipopolysaccharides, resulting

in a negative zeta potential in a neutral medium.22 S. oneidensis is a gram-negative

bacterium that showed a negative zeta potential of –40 mV. S. oneidensis (4.0 × 109 cells)

55

was incubated with sodium formate and tetrachloroauric (III) acid under a nitrogen-

saturated atmosphere at 25°C. After incubation, the precipitate obtained by centrifugation

at 8200 ×g for 10 min was dispersed in ultrapure water.

Figure 4-2. (A) Zeta potential of bacterial suspension during incubation and (B) TEM

images of cells after incubation for (a) 3 and (b) 7 h. (C) TEM images of typical Au NPs

formed on the cells. (a) Seed NPs, (b) sufficient growth Au NPs, and (c) aggregates of Au

NPs.

The zeta potential of S. oneidensis after incubation increased with the incubation time

and became constant at –10 mV over 5 h, as shown Figure 4-2A. In Figure 4-2B, TEM

images show that for Au NPs deposited on the bacterial surface and after incubation for

3 h, a different morphology was observed. In addition to a dominance of Au NPs with a

56

mean diameter of ~2 nm, larger Au NPs with a mean diameter of ~10 nm and their

aggregates were observed on a single cell. Moreover, as the incubation time increased,

the number and size of aggregates of Au NPs increased remarkably on the cell. This

indicates that Au NPs generated as seeds on the cell (seed NPs) continued to grow to 10

nm and aggregated after sufficient growth with increasing incubation time (Figure 4-2C).

This observation supports the results of the spectroscopic measurements. Au NPs were

only observed on cells without free Au NPs on a TEM grid substrate during this process,

suggesting that the production of Au NPs through gateways of extracellularly emitted

electrons, such as MtrC and OmcA, was limited on the bacterial cells. In addition, I found

that Au NPs formed on the cells were covered with a polymeric membrane. Because

bacteria secrete EPS and form biofilms to protect themselves from environmental changes

and chemical substances,23-25 EPS may act as a passivation layer of Au NPs on the cells

and/or reaction field for reducing aurate. Accurate control of the reaction by metabolism

enabled the production of Au NPs with a uniform particle size of 10 nm on the cells.

Therefore, changes in the zeta potential of cells may reflect not only the production of Au

NPs, but also the formation of a biofilm.

S. oneidensis exhibited a weak light-scattering effect based on the difference in the

refractive index between the surrounding air (1.0) and the water (1.3) inside a bacterial

cell, the cytoplasm of which consists of 70% water, 17% proteins, 7% nucleic acids, and

other components (lipids and polysaccharides).17,22,26 The light-scattering intensity of a

single cell increased as the incubation time increased, as shown in Figure. 4-3A,

indicating the reduction of metal ions into (a) Pd and (b) Au NPs by the bacteria.14,15 As

measured at 600 nm, the light-scattering intensity of a single cell increased strongly after

incubation in the suspensions, including the palladium ion, and became constant after 1 h

(Figure 4-3B). After incubation for 1 h, individual bacteria could be clearly observed as

white rods in the dark-field images. According to the results of energy dispersive X-ray

(EDX) spectrometry, it was confirmed that Pd elements were detected on bacterial cells.

Similar phenomena were observed for bacteria incubated in suspensions containing

platinum (Pt) ions. Pt NPs deposited on bacterial cells were observed in dark-field images

after incubation for 3 h. The light-scattering spectrum of bacteria adsorbed with Pt NPs

57

resembled that of bacteria displaying Pd NPs. As measured at 650 nm, the light-scattering

intensity of a single cell incubated in the aurate suspension was observed to gradually

increase during the early stages of incubation and to dramatically increase over 2 h. After

incubation for 4 h, bright, reddish bright rod-like spots, which was attributable to the

formation of Au NPs on the bacteria, could be observed in the dark-field images. After 5

h of incubation, the intensity was two-fold greater than that of palladium ions, although

the formation rate of Pd NPs was higher than that of Au NPs. No changes were observed

in the light-scattering spectra of bacteria incubated in suspensions of other metal ions,

such as copper and nickel. This behavior is attributed to differences in the standard

electrode potential of the metal species and the stability of the metal NPs.

Figure 4-3. (A) Light-scattering spectra of a single cell after incubation in nitrogen-

saturated phosphate buffer supplemented with (a) palladium(II) chloride or (b)

hexachloroauric(III) acid. (B) Dependence of the light-scattering intensity of a single cell

on the incubation time. Dark-field images of a single cell after incubation for 1 and 4 h.

The acquisition time was 400 ms and the scale bar is 2 μm.

58

To investigate whether the species of metal NPs could be determined based on their

light-scattering characteristics, I induced the formation of metal NPs by S. oneidensis in

a mixture of aurate and palladium ions. Bacterial suspensions were incubated at 25℃ in

20 mL of nitrogen-saturated phosphate buffer supplemented with sodium formate (0.10

M), hexachloroauric(III) acid (0.50 mM), and palladium(II) chloride (0.50 mM). After

incubation for 1 h, many whitish rod-like structures were observed in the dark-field

images, indicating the production of Pd NPs on the bacterial cells. The number of reddish

rods increased with passage of the incubation time, and approximately the same number

of whitish and reddish rods were observed after incubation for 4 h, as shown in Figure 4-

4A.

Figure 4-4. (A) Dark-field images and (B) light-scattering spectra of a single cell after

incubation in nitrogen-saturated phosphate buffer supplemented with palladium(II)

chloride and hexachloroauric(III) acid for 4 h. The acquisition time was 400 ms.

59

This suggests that both Au and Pd NPs were produced on the cells. Following an

extension of the incubation time to over 5 h, all bacteria appeared as reddish rods. To

clarify the mechanism underlying this phenomenon, I attempted two-step formation of

metal NPs. S. oneidensis was first incubated in phosphate buffer supplemented with

sodium formate (0.10 M) and palladium (II) chloride (0.50 mM) for 1 h. Pd NP-coated S.

oneidensis was then incubated in phosphate buffer supplemented with sodium formate

(0.10 M) and hexachloroauric(III) acid (0.50 mM). Whereas the Pd signal disappeared

after 4 h of the second incubation, Au was observed in the EDX spectrum. On the contrary,

there was no change in the EDX spectra of S. oneidensis coated with Au NPs before and

after incubation in phosphate buffer supplemented with sodium formate and palladium

(II) chloride (0.50 mM). These results implied that the formation of NPs on the bacterial

surface was controlled by the reduction rate and incubation conditions. 27,28 To apply S.

oneidensis for elemental analysis, a better understanding of the reactivity of metal species

and the stability of metal NPs, based on standard electrode potentials, nanoparticle

forming abilities, and incubation conditions, is required.

4.4. Conclusion

In summary, I evaluated the production of Au NPs on the surface of S. oneidensis by

monitoring their zeta potentials and surface morphologies. The metabolism, including

electron transfer and secretion of EPS, allowed for the reduction of aurate and control of

accurate production of Au NPs with a uniform particle size of 10 nm. I also successfully

identified metal species in a solution based on the light-scattering characteristics of NPs

formed on a bacterial cell. Optimization of the formation of NPs of many metal species,

based on their standard electrode potentials and the incubation conditions, will facilitate

the utilization of S. oneidensis for elemental analysis. Moreover, it is expected that

controlling the incubation conditions of S. oneidensis will enable the application of this

method not only in aqueous solutions, but also in biofilms formed on solid substances.

60

REFERENCES

(1) Ng, C. K.; Tan, T. K. C.; Song, H.; Cao, B. RSC Adv. 2013, 3, 22498.

(2) Suresh, A. K.; Pelletier, D. A.; Wang, W.; Broich, M. L.; Moon, J. W.; Gu, B.; Allison,

D. P.; Joy, D. C.; Phelps, T. J.; Doktycz, M. J. Acta Biomater. 2011, 7, 2148.

(3) Suresh, A. K.; Pelletier, D. A.; Wang, W.; Moon, J. W.; Gu, B.; Mortensen, N. P.;

Allison, D. P.; Joy, D. C.; Phelps, T. J.; Doktycz, M. J. Environ. Sci. Technol. 2010, 44,

5210.

(4) Sheng, L.; Fein, J. B. Environ. Sci. Technol. 2014, 48, 3768.

(5) Qian, F.; Wang, H.; Ling, Y.; Wang, G.; Thelen, M. P.; Li, Y. Nano Lett. 2014, 14,

3688.

(6) Wu, D.; Xing, D.; Lu, L.; Wei, M.; Liu, B.; Ren, N. Bioresouce Technol. 2013, 135,

630.

(7) Tokunou, Y.; Hashimoto, K.; Okamoto, A. J. Phys. Chem. C 2016, 120, 16168.

(8) Yang, Y.; Ding, Y.; Hu, Y.; Cao, B.; Rice, S. A.; Kjelleberg, S.; Song, H. ACS Synth.

Biol. 2015, 4, 815.

(9) Okamoto, A.; Hashimoto, K.; Nakamura, R. Bioelectrochemistry 2012, 85, 61.

(10) Okamoto, A.; Nakamura, R.; Hashimoto, K. Elechtrochim. Acta 2011, 56, 5526.

(11) Bücking, C.; Piepenbrock, A.; Kappler, A.; Gescher, J. Microbiology 2012, 158, 2144.

(12) Morishita, A.; Higashimae, S.; Nomoto, A.; Shiigi, H.; Nagaoka, T. J. Electrochem.

Soc. 2016, 163, G166.

(13) Shi, L.; Squier, T. C.; Zachara, J. M.; Fredrickson, J. K. Mol. Biotechnol. 2007, 65,

12.

(14) Marshall, M. J.; Beliaev, A. S.; Dohnalkova, A. C.; Kennedy, D. W.; Shi, L.; Wang,

Z.; Boyanov, M. I.; Lai, B.; Kemner, K. M.; McLean, J. S.; Reed, S. B.; Culley, D. E.;

61

Bailey, V. L.; Simonson, C. J.; Saffarini, D. A.; Romine, M. F.; Zachara, J. M.;

Fredrickson, J. K. Plos Biol. 2006, 4, 1324.

(15) Ng, C. K.; Sivakumar, K.; Liu, X.; Madhaiyan, M.; Ji, L.; Yang, L.; Tang, C.; Song,

H.; Kjelleberg, S.; Cao, B. Biotechnol. Bioeng. 2013, 110, 1831.

(16) Wu, R.; Cui, L.; Chen, L.; Wang, C.; Cao, C.; Sheng, G.; Yu, H.; Zhao, F. Sci. Rep.

2013, 3, 3307.

(17) Shiigi, H.; Fukuda, M.; Tono, T.; Takada, K.; Okada, T.; Dung, L. Q.; Hatsuoka, Y.

Kinoshita, T.; Takai, M.; Tokonami, S.; Nakao, H.; Nishino, T.; Yamamoto, Y.; Nagaoka,

T. Chem. Commun. 2014, 50, 6252.

(18) Kinoshita, T.; Kiso, K.; Le, D. Q.; Shiigi, H.; Nagaoka, T. Anal. Sci. 2016, 32, 301.

(19) Shiigi, H.; Morita, R.; Yamamoto, Y.; Tokonami, S.; Nakao, H.; Nagaoka, T. Chem.

Commun. 2009, 3615.

(20) Shiigi, H.; Morita, R.; Muranaka, Y.; Tokonami, S.; Yamamoto, Y.; Nakao, H.;

Nagaoka, T. J. Electrochem. Soc. 2012, 159, D442.

(21) Konishi, Y.; Tsukiyama, T.; Tachimi, T.; Saitoh, N.; Nomura, T.; Nagamine, S.

Electrochim. Acta 2007, 53, 186.

(22) Shiigi, H.; Shiigi, T.; Fukuda, M.; Le, D. Q.; Nishino, T.; Nagaoka, T. Anal. Chem.

2015, 87, 4042.

(23) Li, S.; Zhang, X.; Sheng, G. Environ. Sci. Pollut. Res. 2016, 23, 8627.

(24) Li, S.; Sheng, G.; Cheng, Y.; Yu, H. Sci. Rep. 2016, 6, 39098.

(25) Tanzil, A. H.; Sultana, S. T.; Saunders, S. R.; Dohnalkova, A. C.; Shi, L.; Davenport,

E.; Ha, P.; Beyenal, H. Enzyme Microb. Tech. 2016, 95, 69.

(26) Mertens, B.; Blothe, C.; Windey, K.; Windt, W. D.; Verstraete, W. Chemosphere 2007,

66, 99.

(27) Liu, J.; Zheng, Y.; Hong, Z.; Cai, K.; Zhao, F.; Han, H.; Sci. Adv. 2016, 2, e1600858.

62

(28) Corte, S. D.; Hennebel, T.; Fitts, J. P.; Sabbe, T.; Bliznuk, V.; Verschuere, S. D.; Lelie,

van der; Verstraete, W.; Boon, N. Environ. Sci. Technol. 2011, 45 (19), 8506.

63

Chapter V

Kinetics of Intracellular Electron Generation in Shewanella oneidensis MR-1

5.1. Introduction

Bacteria inhabit all biological niches on Earth, and their various activities are

precisely controlled by the flux of electrons, ions, and molecules inside the cell and

between intra- and extracellular environments. To effectively utilize bacteria, it is very

important to quantitatively evaluate their metabolic activity by monitoring

intra/extracellular redox-active substances, such as quinones, nicotinamide adenine

dinucleotide (NADH), and flavin adenine dinucleotide.1−5 Shewanella oneidensis MR-1

is a power-generating bacterium, which possesses an arrangement of cytochromes in the

inner/outer membranes that allows transfer of electrons to an extracellular acceptor, such

as metal ions and/or electrode (Figure 5-1).6−10 Recent studies have indicated that

overexpression of NADH-related genes improved extracellular electron transfer in S.

oneidensis11,12 and that power generation depended on the nature of the carbon source.13,14

The mechanism of switching between ubiquinone and menaquinone was closely related

to the aerobic/anaerobic conversion of the respiratory chain in the cell membrane.15,16

Isotope labeling technique has been widely used for metabolic flux analysis.14,17 This

technique is useful to clarify metabolic pathways based on intracellular contribution of

isotope-labeled organic sources; however, it provides limited kinetic information

concerning intracellular electron generation. These studies improved our understanding

of the intracellular electron transfer and suggested that the intracellular electron

generation was greatly affected by the extracellular electron transfer.

Nevertheless, a more widespread use of this resource has been hampered by the lack

of approaches to precisely quantify bacterial metabolic activity. Conventional techniques,

including chromatography and spectrometry, are complicated assays that consist of pre-

enrichment, centrifugation/filtration, and/or immobilization steps, which hamper

64

obtaining precise kinetic information about metabolic activity.13,17 Therefore, real-time

analytical method for quantitative evaluation of bacterial activity is required.

Figure 5-1. Schematic illustration of formate and lactate metabolism in S. oneidensis.

Although the extracellular electron transfer mechanism of S. oneidensis has been

extensively studied, no detailed qualitative and quantitative assessments of intracellular

electron generation have been conducted. To efficiently utilize S. oneidensis as a

bioresource for applications such as the recovery of metals, preparation of microbial fuel

cells, and treatment of industrial wastewater, it is important to quantitatively evaluate

intracellular electron generation based on individual enzyme reactions in the metabolic

system.

In this section, I have evaluated intracellular electron generation by S. oneidensis,

focusing on the ratio of oxidized and reduced forms of ferricyanide, which are often used

as electron mediator and acceptor.18−20 Real-time measurements by potentiometry in

bacterial suspensions enabled precise quantification of the number of electrons generated

by S. oneidensis based on the Nernst equation, because the [ferricyanide]/[ferrocyanide]

ratio immediately changed during the incubation.

65

5.2. Experimental

5.2.1. Bacterial cultivation

All chemicals (potassium ferricyanide, sodium dihydrogen phosphate dihydrate,

potassium dihydrogen phosphate, sodium formate, sodium lactate, sodium pyruvate, and

acetyl coenzyme A trilithium salt) were of reagent grade and purchased from Wako Pure

Chemical Industries, Ltd. (Tokyo, Japan). Ultrapure water (>18 MΩ cm) sterilized by

ultraviolet light was used for all experiments. The bacterial strain Shewanella oneidensis

MR-1 was purchased from American Type Culture Collection (Manassas, VA, USA).

SYTO 9 green fluorescent nucleic acid stain was purchased from ThermoFisher Scientific.

S. oneidensis was cultured in E-MC35 agar growth medium (Eiken Chemical Co., Tokyo,

Japan) at 303 K for 24 h. Colonies were suspended in liquid nutrient broth (Eiken

Chemical Co.) and incubated at 303 K for 24 h while shaking. After cultivation, the cells

were obtained as a precipitate following centrifugation at 8200g for 10 min, followed by

the resuspension in fresh phosphate-buffered saline (pH 7.4). This procedure was repeated

three times to obtain purified cells. The resulting suspension (6.0 × 108 cells mL–1) was

dispersed in phosphate buffer (pH 7.0) and used for the following experiments. Bacterial

viability was evaluated by counting stained cells with a fluorescent microscope, according

to the manufacturer’s instructions for the BacLight bacterial viability kit (ThermoFisher

Scientific), which consists of two fluorescent pigments, SYTO9 stain for living and dead

cells and propidium iodide stain for dead bacteria.

5.2.2. Potentiometry in bacterial suspensions.

All microbial experiments were performed under strictly sterile conditions. An

Ag|AgCl|saturated KCl| electrode and a gold wire (φ 0.30 mm) were used as reference

and indicator electrodes, respectively. Potentiometry was performed as follows: two

electrodes were immersed in 10 mL of the purified bacterial suspension that included an

organic source (e.g., formate, lactate pyruvate, or acetyl-CoA). Oxygen was removed by

bubbling the suspension with nitrogen (99.99%) for 5 min before ferricyanide addition.

The potential was recorded with an ALS CHi842B Electrochemical Analyzer at ambient

66

temperature (25 ± 1 °C) under nitrogen atmosphere while stirring at 500 rpm. The time at

which ferricyanide reduction by S. oneidensis was complete was determined from the

equivalent point in the potential profile. Reaction rate was calculated by dividing the

initial concentration of ferricyanide by the reaction terminal time. The terminal was

estimated by the equivalent point that represented the maximum value of the potential

change. Apparent reduction rates were obtained by dividing the reduction rates by cell

numbers in suspensions. For each measurement, the number of enzyme molecules in the

cell was considered uniform because incubation conditions and viability were strictly

controlled.

5.3. Results and Discussion

Formate is transported to the periplasm space of S. oneidensis from the extracellular

environment via the outer membrane (Figure 5-1). Electrons generated by the reaction of

formate with formate dehydrogenase (FDH) drive a redox reaction between quinone and

quinol in quinone pool and are subsequently transferred to c-type cytochrome CymA in

the inner membrane. Electrons produced during anaerobic metabolism are transferred

from the cytochrome complex to an extracellular acceptor.9−12,21

The potential (E) in the bacterial suspension containing 10 mM formate changed from

+0.05 V (vs Ag|AgCl) to +0.30 V immediately after the addition of ferricyanide (Figure

5-2A) and gradually decreased throughout the incubation, reaching the quantity potential

(E°′) of ferricyanide. Then, the potential drastically dropped, indicating a terminal point,

whereas the yellowish color of ferricyanide-containing suspension disappeared. These

observations indicated that ferricyanide was almost completely reduced to ferrocyanide

without affecting cell viability (>99%) during the incubation. No potential change was

observed in other bacterial species with FDH. A previous study suggested that metal

nanoparticles formed on the bacterial surface, whereas no particles were observed in the

extracellular environment.22,23 Therefore, the reducing reaction occurred directly at the

numerous electron outlets formed by the cytochrome complex (Figure 5-2B).21 Although

it is convenient to evaluate the [ferricyanide]/[ferrocyanide] ratio spectroscopically, its

changes are difficult to assess due to suspension turbidity. In addition, potentiometry

67

using ferricyanide makes it possible to measure the electron generation within a few

minutes. Therefore, real-time potentiometric measurement without an additional step is

more favorable compared to spectroscopic assay with centrifugation and/or filtration

steps.

Figure 5-2. (A) Potential profiles in bacterial suspensions (3.0 × 109 cells, 10 mM formate,

pH 7.0, 298 K) of facultative anaerobic S. oneidensis and Escherichia coli. An

Ag|AgCl|saturated KCl electrode and a gold wire were used as reference and indicator

electrodes, respectively. Ferricyanide (0.10 mM) was added into the suspension at t = 0 s.

(B) UV–vis spectra of bacterial suspensions in 10 mL of phosphate buffer (pH 7.0). The

inset represents photographs of bacterial pellets of S. oneidensis and E. coli.

Potentiometry enabled monitoring the ferricyanide/ferrocyanide ratio in the

suspension and revealed strong dependence of the terminal time on cell density (Figure

5-3A). The apparent reaction rate (vformate) of 5.1 × 10–16 M cell–1 s–1 was estimated, as the

reaction rate was directly proportional to the cell number (Figure 5-3B). Naturally, the

terminal time was proportional to the initial ferricyanide concentration (Figure 5-3C), but

the reaction rate was almost constant, indicating that electron transfer was negligibly fast

(Figure 5-3D).18−20 The reaction rate strongly depended on the initial concentration of

formate (Figure 5-4Aa), and a typical enzyme reaction curve was obtained (Figure 5-4Ba).

FDH activity was likely a rate-limiting step of electron generation in S. oneidensis,

because formate transport into the intracellular environment did not affect the terminal

time with and without stirring. The Hanes-Woolf plot allowed determination of kinetic

68

parameters, such as maximum rate (Vmax,formate) and Michaelis–Menten constant

(Km,formate), which were 3.7 μM s–1 and 3.9 mM, respectively. The Km,formate value obtained

in S. oneidensis was comparable to those of isolated FDHs.24

Figure 5-3. Potential profiles in bacterial suspensions supplemented with formate. (A)

Potential profiles were measured in bacterial suspensions containing 6.0 × 108–6.0 ×

109cells (pH 7.0, 10 mM formate, 298 K). (B) Relationship between the number of cells

in the suspension and the reaction rate. (C) Dependence of the potential on the initial

concentration of ferricyanide. (D) Relationship between the concentration of ferricyanide

and the reaction rate.

Lactate was previously utilized as a favorable carbon source, and its metabolic

pathway has been described.9−13,25,26 Two electrons generated in the reaction of lactate

with lactate dehydrogenase (LDH) are transferred through redox reactions of

NAD+/NADH and quinone/quinol into the cytochrome complex (Figure 5-1).

Subsequently, produced pyruvate is oxidized through two catabolic pathways comprising

pyruvate dehydrogenase (PDH) and pyruvate formate lyase (PFL). PDH delivers two

electrons obtained by reacting pyruvate with NAD+ yielding acetyl-CoA.27 Acetyl-CoA

69

is partially transformed to acetate to synthesize adenosine triphosphate (ATP). The

remaining acetyl-CoA enters the tricarboxylic acid (TCA) cycle.

Figure 5-4. (A) Potential profiles in bacterial suspensions supplemented with (a) formate

and (b) lactate. Dependence of the potential on the initial concentration of carbon source.

(B) The relationship between the concentration of (a) formate and (b) lactate in the

suspension and reaction rate.

Adding lactate into the suspension changed the potential by S. oneidensis metabolic

system (Figure 5-4Ab). The reaction rate was directly proportional to the cell number,

and the obtained apparent reaction rate (vlactate), 1.9 × 10–16 M cell–1 s–1, was lower than

that obtained with formate (Figure 5-5). The strong dependence of the reaction rate on the

concentration and carbon source species indicated that intracellular enzyme activity was

the rate-limiting step (Figure 5-4Bb). Lactate initially was considered a more preferable

substrate than formate because Vmax,lactate was lower than Vmax,formate, but Km,lactate was 30-

fold lower than Km,formate (0.13 mM of Km,lactate and 1.1 μM s–1 of Vmax,lactate). Moreover, I

found that the reaction rate after the supplementations with both lactate and formate

(vformate+lactate) was nearly equal to the sum of vformate and vlactate (Figure 5-6). This result

indicated that (i) these electron generation pathways were simultaneously active, and

70

subsequent electron transfers were nearly instantaneous, and (ii) the redox processes of

formate-related enzyme (FDH) and lactate-related enzymes (LDH, PFL/PDH, and TCA

cycles) were respectively independent.

Figure 5-5. (A) Potential profiles were measured in bacterial suspensions containing 6.0

× 108–6.0 × 109 cells (pH 7.0, 10 mM lactate, 298 K). (B) Relationship between the

number of cells in the suspension and the reaction rate.

Figure 5-6. Potential profiles in the suspensions supplemented with formate, lactate, or

both formate and lactate.

Potentiometry measures responses rapidly and enables kinetic analysis of electron

production by S. oneidensis during the lag phase in the growth curve.28 Despite lactate

metabolism involves multiple enzyme reactions, repeated additions of ferricyanide

enabled obtaining a constant reduction termination in the suspension that included

sufficient lactate (Figure 5-7Aa). At low lactate concentration, reduction terminal was

delayed drastically by further ferricyanide addition (Figure 5-7Ab). Although the high

71

binding affinity allowed efficient reaction, excessive amount of ferricyanide obviously

delayed it (Figure 5-7B). These results suggested that almost all lactate in the suspension

was converted to pyruvate and, further, to acetyl-CoA, and the contributions of pyruvate

and/or acetyl-CoA to electron generation were much smaller than those of lactate.

Figure 5-7. (A) Potential profiles in the suspension supplemented with (a) 50 mM lactate

and (b) 0.20 mM lactate. Ferricyanide (0.10 mM) was repeatedly supplied into the

suspension. (B) The relationship between the concentration of ferricyanide and the

reaction rate.

Figure 5-8. (A) Potential profiles in the suspensions (6.3 × 109 cells, 0.10 mM

ferrocyanide) supplemented with (a) pyruvate and (b) acetyl-CoA. (B) The relationships

between the concentration of pyruvate and acetyl-CoA and the reaction rate.

Typical potential profiles were obtained in the suspension with pyruvate (Figure 5-

8Aa). The reaction rate (vpyruvate) was 3.5-fold lower than that of vlactate (Figure 5-9A).

72

Similarly, the concentration-potential dependence of acetyl-CoA (Figure 5-8Ab) and a

typical relationship between substrate concentration and reaction rate were observed

(Figure 5-8B). The reaction rate (vacetyl-CoA) was 4.8-fold lower than that of vlactate (Figure

5-9B). Therefore, the reaction rate strongly depended on LDH enzymatic activity due to

the high affinity of LDH binding with substrate.

Figure 5-9. Potential profiles were measured in bacterial suspensions containing 6.0 ×

108–6.0 × 109 cells (A) (pH 7.0, 10 mM pyruvate, 298 K) or (B) (pH 7.0, 10 mM acetyl-

CoA, 298 K), and relationship between the number of cells in the suspension and the

reaction rate.

To evaluate the intracellular electron generation, potentiometric measurements were

performed in the substrate-saturated environment. A uniform potential profile was

obtained in bacterial suspensions with sufficient formate concentrations after 10 additions

of ferricyanide (Figure 5-10Aa). The reaction rate comprised 5.3 × 10–16 ± 6.0 × 10–18 M

cell–1 s–1 (mean ± standard deviation; Table 1), indicating an average of apparent reduction

rates obtained during repeated additions of ferricyanide. This effective reaction rate

(ve_formate) agreed with the apparent vformate value. The electron generation rate (v’formate) by

formate comprised 1.1 × 10–15 M cell–1 s–1, corresponding to 1.1 × 10–12 A cell–1 s–1 of the

73

power generation rate due to the generation of two electrons after the reaction of formate

with FDH.

Figure 5-10. (A) Potential profiles in bacterial suspensions (6.0 × 109 cells, pH 7.0, 298

K) supplemented with (a) formate, (b) lactate, (c) pyruvate, and (d) acetyl-CoA at the

concentration of 50 mM each. Ferricyanide was repeatedly supplied at the indicated

concentration into the suspension. (B) The relationship between the concentration of

carbon source and the reaction rate in bacterial suspensions during repeated additions of

ferricyanide. (C) The relationship between the carbon source and the reaction rate.

Similarly, the reaction rates ve_lactate and ve_pyruvate had low variability in the

suspensions with sufficient amounts of lactate (Figure 5-10Ab) and pyruvate (Figure 5-

10Ac), with the electron generation rates v’lactate and v’pyruvate of 4.2 × 10–16 and 1.6 × 10–

16 M cell–1 s–1, respectively, due to the generation of two electrons in corresponding

reactions. Although the contribution of LDH (2.6 × 10–16 M cell–1 s–1 of vLDH) could be

74

estimated by a subtraction of v’pyruvate from v’lactate (Table 5-2), it was difficult to estimate

the effective reaction rate of acetyl-CoA (Figure 5-10Ad).

Table 5-1. Reaction rates in S. oneidensis suspensions supplemented with different

substrates

At any acetyl-CoA concentration, the reaction rates decreased after ferricyanide

addition (Figure 5-11) and converged to a constant value of 0.023 μM s–1 (Figure 5-10B).

In the cases of lactate and pyruvate, the reaction rates did not change during 10

consecutive additions of ferricyanide (Figure 5-10C). If the contribution of acetyl-CoA to

electron generation was greater, the reaction rates would increase gradually. However, my

result suggested that reduction rates followed Michaelis–Menten mechanism, depending

on the concentration of acetyl-CoA in the suspension after a single addition of

ferricyanide (Figure 5-8B). It is well established that most of acetyl-CoA is converted into

acetate via substrate-level phosphorylation, in which ATP is synthesized by phosphate

acetyltransferase (PTA) and acetate kinase (AK) without electron production (Figure 5-

12A).25,26 The difference of reaction rate between initial and later stages suggested that

acetyl-CoA-related pathway was controlled by a switching mechanism. The constant rate

of 0.023 μM s–1 could be regarded as the contribution of the TCA cycle, because the

reaction rate in the suspension with sufficient amounts of citrate, which is a component

of the TCA cycle, also converged to the same value. The contribution of acetyl-CoA and

TCA cycle to electron generation was consistent and quite lower than those of LDH and

PDH/PFL. The effective reaction rate of v’acetyl-CoA of 3.1 × 10–17 M cell–1 s–1 was obtained

Substrates

Reaction rate (M cell−1 s−1)

Power

generation

(A cell−1 s−1)

Contribution

(%) Apparent, v

Effective, Electron

generation, v

ve SD

Formate 5.1 × 10−16 5.3 × 10−16 6.0 × 10−18 1.1 × 10−15 1.0 × 10−12 64

Lactate 1.9 × 10−16 2.1 × 10−16 5.9 × 10−18 4.2 × 10−16 4.1 × 10−13 25

Pyruvate 5.4 × 10−17 8.0 × 10−17 4.0 × 10−18 1.6 × 10−16 1.5 × 10−13 9.6

Acetyl-CoA 4.0 × 10−17 3.8 × 10−18 - 3.1 × 10−17 3.0 × 10−14 1.8

75

in the 8-electron system of the TCA cycle, corresponding to the reaction rate ve_acetyl-CoA

of 3.8 × 10–18 M cell–1 s–1. The contribution of PDH and FDH through PFL to electron

generation was estimated by subtracting of v’acetyl-CoA from v’pyruvate yielding 1.3 × 10–16

M cell–1 s–1 (vPDH/PFL).

Figure 5-11. Potential profiles during repeated ferricyanide additions into bacterial

suspensions supplemented with acetyl-CoA and citrate. Acetyl-CoA concentrations in the

suspensions (6.0 × 109 cells) were: (A) 1.0, (B) 2.5, (C) 5.0 and (D) 10 mM. Ferricyanide

(0.10 mM) was repeatedly added into the suspension. (E) Relationship between the

concentration of acetyl-CoA and the reaction rate in bacterial suspensions (6.0 × 109 cells)

during repeated additions of ferricyanide. (F) Potential profiles in the suspension (6.0×109

cells) supplemented with 50 mM citrate. Ferricyanide (0.10 mM) was repeatedly supplied

into the suspension.

The fractions of electrons generated by LDH, PDH/PFL, and TCA cycle involved in

lactate metabolism were 0.62, 0.31, and 0.074 as calculated by dividing vLDH, vPDH/PFL

and vTCA cycle by v’lactate, respectively (Figure 5-12B). Moreover, it would be possible to

obtain at least 1.6 × 10–12 A s–1 from a single living cell by efficient progression of these

reactions. It is possible that S. oneidensis could gain energy for survival by fermenting

76

pyruvate into lactate in the absence of available electron acceptor.26,29 In my case,

abundant ferricyanide as electron acceptor enables to effectively transform from lactate

into pyruvate through the enzyme reaction of LDH. In addition, the actual rate of

individual enzyme reaction in the lactate-dependent metabolism was obtained by

subtracting the effects of other enzyme reactions and redox processes as much as possible.

The contribution of ideal electron generation in a single living cell by formate (64%) was

larger than the sum of those achieved by metabolism of lactate (25%), pyruvate (9.6%),

or acetyl-CoA (1.8%).

Figure 5-12. (A) Schematic illustration of the electron transfer in the metabolic pathway.

(B) Schematic diagram of the relationship between the reaction rate and electron

generation rate in (a) the formate-dependent process and (b) the lactate-dependent

metabolism.

77

Table 5-2. Effective reaction rates of electron generation-related enzyme reactions

5.4. Conclusion

By using potentiometry, I have successfully clarified the kinetics of power generation

in S. oneidensis. I found that formate was the most effective carbon source for electron

generation and established the respective contributions of different enzymatic reactions

to lactate metabolism. The small contribution of acetyl-CoA and TCA cycle to electron

generation during lactate metabolism was attributable to the switch from the TCA cycle

to the production of ATP through the PTA–PK process. I believe that quantitative

evaluation of individual enzymatic reactions in the intracellular environment is necessary

for effective utilization of bioresources for practical applications, including the efficient

recovery of metals, preparation of state-of-the-art microbial fuel cells, and effective

treatment of industrial wastewater.

Effective reaction Reaction rate

Contributiona (%)

v (M cell−1 s−1)

v′formate vFDH 1.1 × 10−15 -

v′lactate−v′pyruvate vLDH 2.6 × 10−16 62

v′pyruvate−vacetyl-CoA vPDH/PFL 1.3 × 10−16 31

v′acetyl-CoA vTCA cycle 3.1 × 10−17 7.4

78

REFERENCES

(1) Rabinowitz, J. D.; Vacchino, J. F.; Beeson C. H.; McConnell, H. M. J. Am. Chem. Soc.

1998, 120, 2464.

(2) Canelas, A. B.; ten Pierick, A.; Ras, C.; Seifar, R. M.; van Dam, J. C.; van Gulik, W.

M.; Heijnen, J. J. Anal. Chem. 2009, 81, 7379.

(3) Ikeda, H.; Ishikawa, J.; Hanamoto, A.; Shinose, M.; Kikuchi, H.; Shiba, T.; Sakaki,

Y.; Hattori, M.; Ōmura. S. Nat. Biotechnol. 2003, 21, 526.

(4) Douma, R. D.; de Jonge, L. P.; Jonker, C. T. H.; Seifar, R. M.; Heijnen, J. J.; van Gulik,

W. M. Bioeng. 2010, 107, 105.

(5) Ishiki, K.; Nguyen, D. Q.; Morishita, A.; Shiigi, H.; Nagaoka, T. Anal. Chem. 2018,

90, 10903.

(6) Konishi, Y.; Ohno, K.; Saitoh, N.; Nomura, T.; Nagamine, S.; Hishida, H.; Takahashi,

Y; Uruga, T. J. Biotechnol. 2007, 128, 648–653.

(7) Kim, H. J.; Park, H. S.; Hyun, M. S.; Chang, I. S.; Kim, M.; Kim, B. H. Enzyme

Microb. Technol. 2002, 30, 145–152.

(8) Hirose, A.; Kasai, T.; Aoki, M.; Umemura, T.; Watanabe, K.; Kouzuma, A. Nat.

Commun. 2018, 9, 1083.

(9) Fredrickson, J. K.; Romine, M. F.; Beliaev, A. S.; Auchtung, J. M.; Driscoll, M. E.;

Gardner, T. S.; Nealson, K. H.; Osterman, A. L.; Pinchuk, G.; Reed, J. L.; Rodionov, D.

A.; Rodrigues, J. L.; Saffarini, M. D. A.; Serres, M. H.; Spormann, A. M.; Zhulin, I. B.;

Tiedje, J. M.. Nat. Rev. Microbiol. 2008, 6, 592.

(10) Logan, B. E. Nat. Rev. Microbiol. 2009, 7, 375.

(11) Li, F.; Li, Y. -X.; Cao, Y. -X.; Wang, L.; Liu, C.-G.; Shi, L.; Song, H. Nat. Commun.

2018, 9, 3637.

79

(12) Shi, L.; Dong, H.; Reguera, G.; Beyenal, H.; Lu, A.; Liu, J.; Yu, H.-Q.; Fredrickson.

J. K. Nat. Rev. Microbiol. 2016, 14, 651.

(13) Kane, A. L.; Brutinel, E. D.; Joo, H.; Maysonet, R.; VanDrisse, C. M.; Kotloski, N.

J.; Gralnick, J. A. J. Biotechnol. 2016, 198, 1337.

(14) Luo, S.; Guo, W.; Nealson, K. H.; Feng, X.; He Z. Sci. Rep. 2016, 6, 20941.

(15) Le, D. Q.; Morishita, A.; Tokonami, S.; Nishino, T.; Shiigi, H.; Miyake, M.; Nagaoka,

T. Anal. Chem. 2015, 87, 8416.

(16) Morishita, A.; Higashimae, S.; Nomoto, A.; Shiigi, H.; Nagaoka, T. J. Electrochem.

Soc. 2016, 163, G166.

(17) Krijgsveld, J.; Ketting, R. F.; Mahmoudi, T.; Johansen, J.; Ar-tal-Sanz, M.; Verrijzer,

C. P.; Plasterk, R. H. A; Heck, A. J. R. Nat. Biotechnol. 2003, 21, 927.

(18) Cassatt, J. C.; Marini, C. P. Biochemistry 1974, 13, 5324.

(19) Sutin, N.; Christman, D. R. J. Am. Chem. Soc. 1961, 83, 1773.

(20) Ikeda, T.; Kano, K. K. J. Biosci. Bioeng. 2001, 92, 9.

(21) Pitts, K. E.; Dobbin, P. S.; Ramirez, F. R.; Thomson, A. J.; Richardson, D. J.; Seward,

H. E. J. Biol. Chem. 2003, 278, 27758.

(22) Ishiki, K.; Okada, K.; Le, D. Q.; Shiigi, H.; Nagaoka T. Anal. Sci. 2017, 33, 129.

(23) Ishiki, K.; Shiigi, H.; Nagaoka, T. Anal. Sci. 2017, 33, 55.

(24) Guo, Q.; Gakhar, L.; Wickersham, K.; Francis, K.; Kilshtain, A. V.; Major, D. T.;

Cheatum, C. M.; Kohen, A. Biochemistry, 2016, 55, 2760.

(25) Tang, Y. J.; Meadows, A. L.; Kirby, J.; Keasling, J. D. J. Bacteriol. 2007, 189, 894.

(26) Pinchuk, G. E.; Geydebrekht, O. V.; Hill, E. A.; Reed, J. L.; Konopka, A. E.; Beliaev,

A. S.; Fredrickson, J. K. Appl. Environ. Microbiol. 2011, 77, 8234.

80

(27) Zhou, J.; Olson, D. G.; Lanahan, A. A.; Tian, L.; Murphy. S. J.-L.; Lo, J.; Lynd, L.

R. Biotechnol. Biofuels 2015, 8, 138.

(28) Saito, M.; Ishiki, K.; Nguyen, D. Q.; Shiigi, H. Anal. Chem. 2019, 91, 12793.

(29) Simon, G. M.; Behrens, S.; Choo, A. D.; Spormann, A. M. Appl. Environ. Microbiol.

2007, 73, 1153.

81

Chapter VI

Summary

In this study, I performed the electrochemical and optical evaluation of bacterial

activity. Especially, electrochemical techniques are suitable methods for quantification of

bacterial activity, including oxygen respiration and electron generation. I investigated the

various bacterial metabolism, focusing on the intracellular/extracellular redox species,

membrane proteins and electron transfer.

Chapter 1 is introduction. I described about bacteria, color-based analysis of

metabolic activity, and electrochemical techniques in bioanalysis.

Chapter 2 shows the construction of microbial platform based on conducting polymer

for monitoring bacterial activity. Bacterial cells were immobilized by electrochemical

deposition within conducting polymer including PEDOT and PPy, due to negative zeta

potential around bacterial surface. Microscopic observation revealed that PPy matrix

provides a suitable environment for evaluating bacterial growth and biofilm formation.

The conducting PPy film also make it possible to facilitates electrochemical evaluation

of the respiratory activity of bacterial cells by using a custom-made thin-layer electrolytic

cell. I found that facultative anaerobic and aerobic bacteria exhibit similar respiratory

activities under aerobic conditions. These results indicated that the biological functions

of bacteria were not affected by the chemical structure and electrical conductivity of the

matrix.

Chapter 3 demonstrated an electrochemical detection of viable bacterial cells using

MTT that one of the most useful tools for colorimetric analysis for evaluation cell activity.

I found that MTT can be applied to not colorimetric but electrochemical assay. I

successfully deposited insoluble formazan generated by bacterial cells on ITO electrode

by drying suspension. This technique was capable of detecting microbes above 2.8 × 101

CFU mL–1 and required only a 1 h incubation. The sharp oxidation peak based on

formazan oxidation make it possible to quantify viable cell numbers. The results of this

study indicate that the sensitivity of the present technique is up to 10,000-fold higher than

82

that of MTT colorimetry. The higher sensitivity is mainly ascribed to the reduction of

microbially produced formazan to an extremely small volume through evaporation,

coupled with the baseline-separated intense adsorption peak.

Chapter 4 describes metal-ion reduction by S. oneidensis. First, I tracked formation

process of Au NPs on the S. oneidensis cell surface by microscopic techniques, and found

that EPS allowed for the reduction of metal ion and control the size of nanoparticles. The

uniform production of nanoparticle can be applied to optical elemental analysis of the

respective nanoparticles. I successfully identified metal species in solution based on the

light-scattering property of NPs formed on bacterial cells. Controlling the incubation

condition of S. oneidensis make it possible to apply this method not only in aqueous

solution but also in biofilm formed on solid substances.

Chapter 5 denote real-time evaluation of intracellular electron generation in S.

oneidensis. Potentiometry make it possible to quantify of the number of electrons

generated by S. oneidensis based on the Nernst equation. The amount of electron

generation strongly depended on the nature of the carbon source. Analysis of the obtained

kinetic parameters of intracellular electron generation demonstrated that formate was the

most effective carbon source, as it enabled 2.5-fold faster electron generation rate than

other sources. I established that the respective contributions of lactate dehydrogenase,

pyruvate dehydrogenase/pyruvate-formate-lyase, and tricarboxylic acid cycle to lactate

metabolism were 62%, 31%, and 7.4%, correspondingly. Furthermore, I clarified that

electrons may be generated at 1.6 × 10–12 A s–1 by ideal metabolism in a single living cell.

These findings establish the basis for biological strategies of electron production and

facilitate the utilization of S. oneidensis as a bioresource in practical applications,

including energy production, environmental purification, and recovery of useful materials.

Chapter 6 summarized the whole results and conclusions of the thesis.

I developed an electrochemical method for quantitative evaluation of bacterial activity.

Focusing on the electrochemical response of intracelluar/extracellular redox species, I

succeeded in measuring bacterial activity such as respiration, metal-ion reduction, and

83

electron transfer in real-time. I believed that electrochemical techniques will more and

more develop as new methods for quantifying bacterial activity, since there are many

measurement techniques such as voltammetry, potentiometry, or coulometry, and high

resolution of electrochemical devices make it possible to evaluate bacterial metabolism

at single-cell order.

84

ACKNOWLEDGEMENTS

Firstly, I would like to appreciate to my supervisor, Professor Hiroshi Inoue for a lot

of guidance and support for my doctoral course.

I also would like to express my appreciation to Prof. Hideaki Hisamoto and Prof.

Atsushi Harada for their carefully reading of my thesis.

Next, I would like to express my sincere appreciation to Emeritus Professor Tsutomu

Nagaoka and Associate Professor Hiroshi Shiigi. Their polite and sometimes harsh

guidance helped me in all the time of research and writing of this thesis. I also thank to

Yojiro Yamamoto (Green Chem Inc.). He helped me with all aspects of the experiment,

including how to use the laboratory equipment in the lab and how to prepare the

experimental sample. I also appreciate to Ms. Eriko Shimizu who kindly advised for my

financial procedure.

I would like to thank all the wonderful lab members for their exciting discussions.

Special thanks to Dr. Takamasa Kinoshita and Dr. Shan Shueling our laboratory seniors.

They gave me a lot of experimental advice and made my doctoral life enjoyable. Dr.

Nguyen Quang Dung, and Ms. Maki Saito often helped me for experiments and various

things in laboratory life.

I acknowledge financial support from Sasakura Enviro-Science Foundation and the

Japanese Society for the Promotion of Science (JSPS) through a Grant-in-Aid for JSPS

Research Fellow (19J10509).

Finally, I would like to thank to my family for their warm and patient support over

the years.

85

LIST OF PUBLICATIONS

No. Title of the article Authors Journal’s name, Vol., Pages, and Year

Corresponding chapter

1 A Microbial Platform Based on Conducting Polymers for Evaluating Metabolic Activity

M. Saito K. Ishiki D. Q. Nguyen H. Shiigi

Analytical Chemistry, Vol. 91, pp. 12793-12798 (2019).

Chapter 2

2 Electrochemical Detection of Viable Bacterial Cells Using a Tetrazolium Salt

K. Ishiki D. Q. Nguyen A. Morishita H. Shiigi T. Nagaoka

Analytical Chemistry, Vol. 90, pp. 10903-10909 (2018).

Chapter 3

3 Investigation Concerning the Formation Process of Gold Nanoparticles by Shewanella oneidensis MR-1

K. Ishiki K. Okada D. Q. Le H. Shiigi T. Nagaoka

Analytical Sciences, Vol. 33, pp. 129-131 (2017).

Chapter 4

4 Optical Elemental Analysis of Metals Using Shewanella oneidensis

K. Ishiki H. Shiigi T. Nagaoka

Analytical Sciences, Vol. 33, pp. 551-553 (2017).

Chapter 4

5 Kinetics of Intracellular Electron Generation in Shewanella oneidensis MR-1

K. Ishiki H. Shiigi

Analytical Chemistry, Vol. 91, pp. 14401-14406 (2019).

Chapter 5

86

LIST OF OTHER PUBLICATIONS

No. Title of the article Authors Journal’s name, Vol., Pages, and Year

1 Nanoantenna for Bacterial Detection T. Kinoshita M. Fukuda D. Q. Nguyen K. Ishiki T. Nishino H. Shiigi T. Nagaoka

Procedia Chemistry, Vol. 20, pp. 90-92 (2016).

2 Shape Memory Characteristics of O157-Antigenic Cavities Generated on Nanocomposites Consisting of Copolymer-Encapsulated Gold Nanoparticles

T. Kinoshita D. Q. Nguyen D. Q. Le K. Ishiki T. Nishino H. Shiigi T. Nagaoka

Analytical Chemistry, Vol. 89, pp. 4680-4684 (2017).

3 Real-Time Evaluation of Bacterial Viability Using Gold Nanoparticles

T. Kinoshita K. Ishiki D. Q. Nguyen H. Shiigi T. Nagaoka

Analytical Chemistry, Vol. 90, pp. 4098-4103 (2018).

4 Single Cell Immunodetection of Escherichia coli O157:H7 on an Indium-Tin-Oxide Electrode by Using an Electrochemical Label with an Organic-Inorganic Nanostructure

D. Q. Nguyen K. Ishiki H. Shiigi

Microchimica Acta, Vol. 185, pp. 465(1-8) (2018).

5 Smart Golden Leaves Fabricated by Integrating Au Nanoparticles and Cellulose Nanofibers

H. Shiigi T. Tomiyama M. Saito K. Ishiki D. Q. Nguyen T. Endo Y. Yamamoto X. Shan Z. Chen T. Nishino H. Nakao T. Nagaoka

ChemNanoMat, Vol. 5, pp. 581-585 (2019).